Amfanin MemTrax da Samfuran Koyon Na'ura a cikin Rarraba Rawan Fahimci

Rubutun Bincike

Marubuta: Bergeron, Michael F. | Landset, Sara | Zhou, Xianbo | Ding, Tao | Khoshgoftaar, Taghi M. | Zhao, Feng | Du, Bo | Chen, Xinjie | Wang, Xuan | Zhong, Lianmei | Liu, Xiaolei| Ashford, J. Wesson

DOI: 10.3233/JAD-191340

Jaridar: Jaridar Cutar Alzheimer, kundi. 77, a'a. 4, shafi na 1545-1558, 2020

Abstract

Bayan Fage:

Yaɗuwar abin da ya faru da yaɗuwar Alzheimer ta cutar da raunin fahimi mai sauƙi (MCI) ya haifar da kiran gaggawa don bincike don tabbatar da gano farkon ganowa da tantancewa.

Manufa:

Manufar binciken mu na farko shine don tantance idan zaɓaɓɓen ma'aunin aikin MemTrax da ƙididdige ƙididdiga masu dacewa da halayen bayanan kiwon lafiya za a iya amfani da su yadda ya kamata a cikin ƙirar tsinkaya da aka haɓaka tare da koyon injin don rarraba lafiyar fahimi (na al'ada da MCI), kamar yadda za a nuna ta Binciken Nazarin Montreal (MoCA).

Hanyar:

Mun gudanar da wani bincike-bincike akan 259 neurology, asibitin ƙwaƙwalwar ajiya, da kuma tsofaffi marasa lafiya na ciki da aka dauka daga biyu. asibitoci a kasar Sin. An bai wa kowane majiyyaci MoCA na yaren Sinanci kuma ya gudanar da kansa na ci gaba da sanin MemTrax akan layi. gwajin ƙwaƙwalwar ajiya akan layi a rana guda. An gina ƙididdigan ƙididdiga ta amfani da koyan inji tare da ingancin giciye mai ninki 10, kuma an auna aikin ƙirar ta amfani da Wurin Ƙarƙashin Ƙarƙashin Ƙarƙashin Ƙarƙashin Mai karɓa (AUC). An gina samfura ta amfani da ma'aunin aikin MemTrax guda biyu (kashi daidai, lokacin amsawa), tare da fasalulluka takwas na gama-gari da tarihin sirri.

results:

Kwatanta xaliban a cikin zaɓaɓɓun haɗaɗɗun maki na MoCA da ƙofa, Naïve Bayes gabaɗaya shi ne babban koyi mai ƙwazo tare da rarrabuwa gabaɗaya na 0.9093. Bugu da ari, a cikin manyan xalibai uku, aikin rarrabuwa na tushen MemTrax gabaɗaya ya kasance mafifici ta amfani da manyan siffofi huɗu (0.9119) idan aka kwatanta da amfani da duk fasalulluka 10 gama gari (0.8999).

Kammalawa:

Ana iya amfani da aikin MemTrax yadda ya kamata a cikin ƙirar ƙira na koyo na inji aikace-aikacen nunawa don gano rashin lafiyar matakin farko.

GABATARWA

Abubuwan da aka sani (duk da cewa ba a gano su ba) yaduwa da yaduwa da yaɗuwa da haɓakar likita, zamantakewa, da jama'a. kiwon lafiya farashi da nauyin cutar Alzheimer (AD) da ƙarancin fahimi (MCI) suna ƙara wahala ga duk masu ruwa da tsaki [1, 2]. Wannan lamari mai ban tsoro da ban tsoro ya haifar da kiran gaggawa don bincike don tabbatarwa ganowa da wuri kayan aikin tantance fahimi da kayan kima don amfanin yau da kullun na yau da kullun a cikin sirri da saitunan asibiti don tsofaffin marasa lafiya a yankuna daban-daban da yawan jama'a [3]. Waɗannan kayan aikin dole ne kuma su samar da fassarorin sakamako marasa kyau zuwa bayanan lafiyar lantarki. Za a sami fa'idodin ta hanyar sanar da marasa lafiya da kuma taimaka wa likitoci don gane manyan canje-canje a baya kuma don haka ba da damar haɓaka da sauri da daidaitawa akan lokaci, aiwatarwa, da bin diddigin daidaitattun daidaikun mutane da ƙarin farashi mai tsada da kulawa da haƙuri ga waɗanda suka fara dandana. raguwar fahimi [3, 4].

Kayan aikin MemTrax na kwamfuta (https://memtrax.com) ƙima ce mai sauƙi da ɗan taƙaitaccen ci gaba da ƙima wanda za'a iya sarrafa kansa akan layi don auna ƙalubalen aikin ƙwaƙwalwar lokaci na lokaci inda mai amfani ya amsa maimaita hotuna ba ga gabatarwar farko ba [5, 6]. Binciken da aka yi kwanan nan da abubuwan da suka haifar da aiki sun fara ci gaba da ci gaba kuma tare da nuna ingancin asibiti na MemTrax a farkon AD da MCI nunawa [5-7]. Koyaya, kwatanta kai tsaye na amfanin asibiti zuwa data kasance lafiyar hankali kimantawa da ma'auni na al'ada suna da garanti don sanar da hangen nesa na ƙwararru da kuma tabbatar da amfanin MemTrax a farkon ganowa da tallafin bincike. van der Hoek et al. [8] idan aka kwatanta zaɓaɓɓun ma'aunin aikin MemTrax (gudun amsa da daidai daidai) zuwa matsayin fahimi kamar yadda Montreal ta ƙaddara. Ƙimar Fahimta (MoCA). Duk da haka, wannan binciken ya iyakance ga haɗa waɗannan ma'auni na aiki tare da halayyar halin fahimi (kamar yadda MoCA ya ƙaddara) da ma'anar kewayon dangi da ƙimar yankewa. Don haka, don faɗaɗa kan wannan bincike da haɓaka aikin rarrabuwa da inganci, tambayar bincikenmu ta farko ita ce:

• Za a iya zaɓaɓɓen ma'aunin aikin MemTrax da ƙididdiga masu dacewa da lafiya Cikakken Bayani za a yi amfani da halaye yadda ya kamata a cikin ƙirar tsinkaya da aka ƙera tare da koyan injin don rarraba lafiyar fahimi a hankali (na al'ada da MCI), kamar yadda makin MoCA ɗin mutum zai nuna?

Na biyu ga wannan, muna so mu sani:

• Ciki har da fasali iri ɗaya, shin za a iya amfani da samfurin koyon aikin injin na MemTrax da kyau ga majiyyaci don hasashen tsanani (mai sauƙi da mai tsanani) a cikin zaɓaɓɓun nau'ikan nakasawar fahimi kamar yadda bincike na asibiti mai zaman kansa zai ƙaddara?

Zuwan da haɓaka aikace-aikacen fasaha na wucin gadi da koyan injina a cikin dubawa / ganowa sun riga sun nuna fa'idodi masu amfani daban-daban, tare da ƙirar tsinkaya yadda ya kamata ke jagorantar likitocin a cikin ƙalubale na kimanta fahimi / lafiyar kwakwalwa da sarrafa haƙuri. A cikin bincikenmu, mun zaɓi irin wannan hanya a cikin ƙirar ƙira ta MCI da nuna bambanci mai tsanani na rashin fahimta kamar yadda aka tabbatar ta hanyar bincike na asibiti daga ɗakunan bayanai guda uku waɗanda ke wakiltar zaɓaɓɓun marasa lafiya da marasa lafiya daga asibitoci biyu a China. Ta yin amfani da ƙirar ƙira na koyan inji, mun gano ƙwararrun ƙwararrun ƙwararrun ƙwararrun ƙwararrun ƙwararrun ƙwararrun bayanai/ haɗaɗɗiyar masu koyo kuma mun sanya fasalulluka don jagorance mu wajen ayyana mafi kyawun aikace-aikacen ƙirar ƙirar asibiti.

Hasashenmu shine cewa za'a iya amfani da ingantaccen samfurin tushen MemTrax don rarraba lafiyar fahimi kai tsaye (na al'ada ko MCI) bisa ma'aunin ma'aunin ƙimar MoCA, kuma ana iya amfani da samfurin tsinkayar MemTrax mai kama da inganci wajen nuna wariya a cikin zaɓaɓɓun nau'ikan. asibiti bincike rashin fahimta. Nuna sakamakon da ake tsammani zai zama kayan aiki don tallafawa ingancin MemTrax azaman allon ganowa na farko don raguwar fahimi da rabe-raben fahimi. Ingantacciyar kwatancen masana'antu da aka ce ma'aunin masana'antu wanda ya dace da sauƙin sauƙi da saurin amfani zai yi tasiri a cikin taimaka wa likitocin su ɗauki wannan kayan aiki mai sauƙi, abin dogaro, da samun dama a matsayin allo na farko a gano farkon (ciki har da prodromal) gazawar fahimi. Irin wannan hanya da amfani na iya haifar da ƙarin lokaci kuma mafi kyawun kulawar majiyyaci da sa baki. Wadannan hangen nesa na gaba da ingantattun ma'auni da ƙira na iya zama taimako wajen ragewa ko dakatar da ci gaban cutar hauka, gami da AD da kuma AD-related dementias (ADRD).

KAYA DA MATAKAI

Masanin binciken

Tsakanin Janairu 2018 da Agusta 2019, an kammala bincike-bincike a kan marasa lafiya da aka dauka daga asibitoci biyu a kasar Sin. Gudanar da MemTrax [5] ga mutanen da ke da shekaru 21 zuwa sama kuma tattarawa da nazarin waɗannan bayanan an sake duba su kuma an amince da su kuma an gudanar da su daidai da ƙa'idodin ɗabi'a na Human Kwamitin Kare Magana na Jami'ar Stanford. MemTrax da duk sauran gwaje-gwaje na wannan binciken gabaɗaya an yi su ne bisa ga sanarwar Helsinki na 1975 kuma kwamitin nazarin cibiyoyi na Asibitin Farko na Jami'ar Kiwon Lafiya ta Kunming da ke Kunming, Yunnan, China ta amince. An ba kowane mai amfani da wani sanar da izini fom don karantawa/bita sannan da yardan shiga da son rai.

An dauki mahalarta daga tafkin na marasa lafiya a asibitin jijiyoyi a asibitin Yanhua (YH sub-dataset) da kuma asibitin ƙwaƙwalwar ajiya a Asibitin Haɗin Kan Farko na Kunming Medical Jami'ar (XL sub-dataset) a Beijing, China. An kuma dauki mahalarta daga ilimin jijiya (XL sub-dataset) da kuma na cikin gida (KM sub-dataset) marasa lafiya a Asibitin Farko na Jami'ar Kiwon Lafiya ta Kunming. Sharuɗɗan haɗawa sun haɗa da 1) maza da mata aƙalla ’yan shekara 21, 2) ikon jin Sinanci (Mandarin), da 3) ikon fahimtar kwatance na magana da rubutu. Sharuɗɗan keɓancewa sune hangen nesa da nakasar motsa jiki da ke hana mahalarta kammalawa Gwajin MemTrax, kazalika da rashin iya fahimtar takamaiman umarnin gwaji.

Sigar Sinanci na MemTrax

A kan layi An fassara dandalin gwajin MemTrax zuwa Sinanci (URL: https://www.memtrax.com.cn) kuma an ƙara daidaita shi don amfani da shi ta hanyar WeChat (Shenzhen Tencent Computer Systems Co. LTD., Shenzhen, Guangdong, China) don gudanar da kai. An adana bayanai a kan uwar garken girgije (Ali Cloud) da ke kasar Sin kuma an ba da lasisi daga Alibaba (Alibaba Technology Co. Ltd., Hangzhou, Zhejiang, China) ta SJN Biomed LTD (Kunming, Yunnan, China). Takamaiman cikakkun bayanai akan MemTrax da ka'idojin ingancin gwajin da aka yi amfani da su anan an bayyana su a baya [6]. An bayar da gwajin ba tare da caji ba ga marasa lafiya.

Tsarin karatu

Ga marasa lafiya da marasa lafiya, takardar tambayoyin gama gari don tattara bayanan jama'a da na sirri kamar shekaru, jima'i, shekarun ilimi, sana'a, rai kadai ko tare da dangi, kuma memba na ƙungiyar binciken ya gudanar da tarihin likita. Bayan kammala tambayoyin, an gudanar da gwajin MoCA [12] da MemTrax (MoCA farko) ba tare da fiye da mintuna 20 tsakanin gwaje-gwaje ba. MemTrax bisa dari daidai (MTx-% C), ma'anar lokacin amsawa (MTx-RT), da kwanan wata da lokacin gwajin an rubuta su a takarda ta wani memba na ƙungiyar binciken don kowane ɗan takara da aka gwada. An ɗora tambayoyin da aka kammala da sakamakon MoCA a cikin ma'auni na Excel ta mai binciken wanda ya gudanar da gwaje-gwajen kuma abokin aiki ya tabbatar da shi kafin a adana fayilolin Excel don nazari.

Gwajin MemTrax

Gwajin kan layi na MemTrax ya haɗa da hotuna 50 (25 na musamman da maimaita 25; saiti 5 na hotuna 5 na al'amuran gama gari ko abubuwa) waɗanda aka nuna a cikin takamaiman tsari na bazuwar. Mahalarcin zai (a kowane umarni) ya taɓa maɓallin farawa akan allon don fara gwajin kuma ya fara duba jerin hotunan sannan ya sake taɓa hoton akan allon da sauri a duk lokacin da maimaita hoto ya bayyana. Kowane hoto ya bayyana tsawon s 3 ko har sai an taɓa hoton da ke kan allon, wanda ya haifar da gabatar da hoto na gaba nan da nan. Yin amfani da agogon ciki na na'urar gida, MTx-RT na kowane hoto an ƙaddara ta lokacin da ya wuce daga gabatar da hoton zuwa lokacin da ɗan takara ya taɓa allon don nuna alamar gane hoton a matsayin wanda aka riga aka nuna. a lokacin gwaji. An yi rikodin MTx-RT don kowane hoto, tare da cikakken rikodin 3 s wanda ke nuna babu amsa. An ƙididdige MTx-% C don nuna adadin maimaitawa da hotuna na farko waɗanda mai amfani ya amsa daidai (gaskiya tabbatacce + korau na gaskiya ya raba ta 50). Ƙarin cikakkun bayanai na gudanarwa da aiwatarwa na MemTrax, raguwar bayanai, bayanan mara inganci ko "babu amsa", da kuma bayanan bayanan farko an bayyana su a wani wuri [6].

An yi bayanin gwajin MemTrax daki-daki kuma an ba da gwajin gwaji (tare da hotuna na musamman ban da waɗanda aka yi amfani da su a cikin gwajin don rikodin sakamakon) ga mahalarta a cikin saitunan asibiti. Masu shiga cikin ƙananan bayanan YH da KM sun ɗauki gwajin MemTrax a kan wayar salula wanda aka ɗora tare da aikace-aikacen akan WeChat; yayin da iyakataccen adadin majinyatan bayanan bayanan XL sun yi amfani da iPad yayin da sauran suka yi amfani da wayar hannu. Duk mahalarta sun ɗauki gwajin MemTrax tare da mai binciken binciken ba tare da damuwa ba.

Ƙimar fahimi na Montreal

Sigar Beijing ta MoCA ta Sinawa (MoCA-BC) [13] an gudanar da ita kuma ta samu nasara daga kwararrun masu bincike bisa ga umarnin gwaji na hukuma. Da kyau, an nuna MoCA-BC ya zama abin dogaro gwajin fahimi An yi nazari a duk matakan ilimi a cikin tsofaffi na kasar Sin [14]. Kowane gwaji ya ɗauki kimanin mintuna 10 zuwa 30 don gudanarwa bisa la'akari da iyawar ɗan takara.

MoCA rarrabuwa yin tallan kayan kawa

Akwai jimillar fasalulluka 29 masu amfani, gami da MemTrax guda biyu gwada ma'aunin aiki da fasali 27 masu alaƙa da alƙaluma da lafiya bayanai ga kowane ɗan takara. An yi amfani da makin gwajin jimlar MoCA na kowane majiyyaci azaman fahimi nunawa "ma'auni" don horar da samfuran tsinkaya. Saboda haka, saboda an yi amfani da MoCA don ƙirƙirar lakabin aji, ba za mu iya amfani da jimillar makin (ko kowane makin na MoCA) azaman siffa mai zaman kanta ba. Mun yi gwaje-gwaje na farko wanda a ciki muka tsara (raba lafiyar fahimi da MoCA ta ayyana) ainihin ƙananan bayanan asibiti / asibiti (s) guda uku daban-daban sannan a haɗa su ta amfani da duk fasalulluka. Koyaya, ba a tattara duk abubuwan bayanan iri ɗaya ba a cikin kowane ɗayan asibitocin guda huɗu waɗanda ke wakiltar ƙananan bayanan uku; don haka, yawancin fasalullukan mu a cikin haɗaɗɗun bayanan bayanai (lokacin da ake amfani da duk fasalulluka) sun sami babban abin da ya ɓace na ƙimar ƙima. Daga nan mun gina samfura tare da haɗaɗɗiyar saitin bayanai ta amfani da fasalulluka gama-gari waɗanda suka haifar da ingantacciyar aikin rarrabuwa. Wataƙila an bayyana wannan ta hanyar haɗuwa da samun ƙarin lokuta don yin aiki tare da ta hanyar haɗa ƙananan bayanan bayanan marasa lafiya guda uku kuma babu wani fasali tare da ƙarancin ƙimar ƙimar da ba ta dace ba (siffa ɗaya kawai a cikin haɗin bayanan da aka haɗa, nau'in aiki, yana da ƙima mai ɓacewa, yana tasiri. lokuta uku ne kawai na haƙuri), saboda kawai abubuwan gama gari da aka rubuta a duk rukunin yanar gizon uku an haɗa su. Musamman ma, ba mu da takamaiman ma'aunin ƙi ga kowane fasalin da a ƙarshe ba a haɗa shi a cikin haɗaɗɗiyar bayanai ba. Koyaya, a cikin tsarin ƙirar bayananmu na farko na haɗe-haɗe, mun fara amfani da duk fasalulluka daga kowane rukunin bayanan marasa lafiya guda uku daban-daban. Wannan ya haifar da aikin ƙirar ƙira wanda ya yi ƙasa da ƙima fiye da ƙirar farko ta farko akan kowane ƙaramin bayanai. Bugu da ƙari, yayin da aikin rarrabuwar samfuran da aka gina ta amfani da duk fasalulluka ya kasance mai ƙarfafawa, a cikin dukkan ɗalibai da tsare-tsaren rarrabuwa, aikin ya inganta sau biyu fiye da ƙira yayin amfani da abubuwan gama gari kawai. A haƙiƙa, a cikin abin da ya ƙare zama ƙwararrun ɗalibanmu, duk abin da ya inganta sai dai ɗaya samfurin ya kawar da abubuwan da ba na kowa ba.

Ƙididdigar ƙididdiga ta ƙarshe (YH, XL, da KM hade) sun haɗa da lokuta 259, kowannensu yana wakiltar ɗan takara na musamman wanda ya ɗauki duka MemTrax da gwaje-gwajen MoCA. Akwai siffofi masu zaman kansu guda 10 da aka raba: MemTrax ma'aunin aikin aiki: MTx-% C da ma'anar MTx-RT; Bayanin tarihin alƙaluma da tarihin likita: shekaru, jima'i, shekarun ilimi, nau'in aiki (ƙwanƙwasa shuɗi / farar kwala), tallafin zamantakewa (ko mai gwadawa yana zaune shi kaɗai ko tare da dangi), da e/a'a amsoshin ko mai amfani yana da tarihin ciwon sukari, hyperlipidemia, ko raunin kwakwalwa. An yi amfani da ƙarin ma'auni guda biyu, ma'auni na MoCA da maki na MoCA wanda aka daidaita don shekaru na ilimi [12], an yi amfani da su daban don haɓaka alamun rarrabuwa masu dogaro, don haka ƙirƙirar ƙirar ƙira guda biyu daban-daban da za a yi amfani da su a haɗin bayanan mu. Ga kowane juzu'i (daidaitacce da ba a daidaita su) na makin MoCA, an sake keɓance bayanan daban don rarrabuwar binaryar ta amfani da ma'auni daban-daban guda biyu-wanda aka ba da shawarar farko [12] da madadin ƙimar da wasu ke amfani da su kuma suka haɓaka [8, 15]. A cikin madaidaicin tsarin rarraba kofa, an yi la'akari da majiyyaci yana da lafiyar fahimi na yau da kullun idan ya / ya ci ≥23 akan gwajin MoCA kuma yana da MCI idan maki ya kasance 22 ko ƙasa; yayin da, a cikin tsarin rarrabuwa na farko da aka ba da shawarar, mai haƙuri ya sami maki 26 ko mafi kyau akan MoCA don a lakafta shi azaman lafiyar fahimi na yau da kullun.

Bayanan da aka tace don ƙirar ƙira ta MoCA

Mun ƙara yin nazarin rarrabuwar MoCA ta amfani da dabarun martaba guda huɗu da aka saba amfani da su: Chi-Squared, Ratio Ratio, Samun Bayanai, da Rashin tabbas. Don hangen nesa na wucin gadi, mun yi amfani da masu matsayi zuwa ga dukkan bayanan da aka haɗa ta amfani da kowane tsarin ƙirar mu guda huɗu. Duk masu daraja sun yarda akan manyan siffofi iri ɗaya, watau, shekaru, adadin shekarun ilimi, da duka ma'aunin aikin MemTrax (MTx-% C, yana nufin MTx-RT). Sannan mun sake gina samfuran ta amfani da kowane dabarar zaɓin fasalin don horar da ƙirar akan manyan abubuwa huɗu kawai (duba Zaɓin fasalin kasa).

Sakamakon bambance-bambancen takwas na ƙarshe na tsare-tsaren rarrabuwar maki na MoCA an gabatar da su a cikin Tebur 1.

Tebur 1

Takaitacciyar bambance-bambancen tsarin ƙirar ƙira da aka yi amfani da su don rarraba MoCA (Normal Lafiyar Fahimi ya da MCI)

Tsarin Samfura	Kiwon Lafiyar Fahimi na Al'ada (Aji mara kyau)	MCI (Aji mai kyau)
Daidaita-23 Ba a Tace/Ba a Tace ba	101 (39.0%)	158 (61.0%)
Daidaita-26 Ba a Tace/Ba a Tace ba	49 (18.9%)	210 (81.1%)
Ba a daidaita ba-23 Ba a tace/Ba a tace ba	92 (35.5%)	167 (64.5%)
Ba a daidaita ba-26 Ba a tace/Ba a tace ba	42 (16.2%)	217 (83.8%)

Adadi da kashi dari na jimlar marasa lafiya a cikin kowane aji an bambanta su ta hanyar daidaita ma'auni don ilimi (daidaitacce ko ba a daidaita ba) da matakin ƙima (23 ko 26), kamar yadda aka yi amfani da su ga saitin fasali guda biyu (Ba a tacewa da Filtered).

Samfuran ƙima na tushen asibiti na MemTrax

Daga cikin bayanan mu na asali guda uku (YH, XL, KM), kawai majinyatan bayanan bayanan XL ne kawai aka bincikar su ta asibiti don rashin fahimi (watau, ba a yi amfani da makin MoCA na su ba wajen kafa rarrabuwa na al'ada da nakasa). Musamman, an gano marasa lafiyar XL tare da ko dai Gwajin cutar Alzheimer (AD) ko ciwon daji na jijiyoyin jini (VaD). A cikin kowane ɗayan waɗannan nau'ikan ganewar asali na farko, an sami ƙarin ƙira don MCI. Bincike na MCI, lalata, rashin lafiyar neurocognitive na jijiyoyi, da kuma rashin lafiyar neurocognitive saboda AD sun dogara ne akan ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙayyadaddun ƙwayar cuta: DSM-5 [16]. Idan aka yi la'akari da waɗannan ƙayyadaddun ƙayyadaddun cututtukan, an yi amfani da tsarin ƙirar ƙira guda biyu daban-daban zuwa ƙananan bayanan XL don bambanta matakin tsanani (matakin naƙasa) ga kowane nau'in ganewar asali na farko. Bayanan da aka yi amfani da su a cikin kowane ɗayan waɗannan tsare-tsaren ƙirar ƙira (AD da VaD) sun haɗa da bayanan alƙaluma da tarihin haƙuri, da kuma aikin MemTrax (MTx-% C, yana nufin MTx-RT). Kowane ganewar asali an lakafta shi mai laushi idan an tsara MCI; in ba haka ba, an dauke shi mai tsanani. Da farko mun yi la'akari da haɗawa da ƙimar MoCA a cikin samfuran ganewar asali (mai laushi da mai tsanani); amma mun yanke shawarar hakan zai karya manufar tsarin ƙirar mu na tsinkaya na sakandare. Anan za a horar da xaliban ta hanyar amfani da wasu halaye masu haƙuri waɗanda ke samuwa ga mai bayarwa da awoyi na gwajin MemTrax mafi sauƙi (a madadin MoCA) a kan ma'anar "ma'auni na zinariya", ganewar asibiti mai zaman kanta. Akwai lokuta 69 a cikin bayanan bincike na AD da kuma lokuta 76 na VaD (Table 2). A cikin duka bayanan, akwai fasali masu zaman kansu guda 12. Baya ga siffofi 10 da aka haɗa a cikin ƙididdigar ƙima na MoCA, tarihin haƙuri ya haɗa da bayani game da tarihin hauhawar jini da bugun jini.

Tebur 2

Takaitacciyar bambance-bambancen tsarin ƙirar ƙirar ƙira da aka yi amfani da shi don rarrabuwa mai tsanani (Mild vs Severe)

Tsarin Samfura	M (Aji mara kyau)	Tsanani (Aji Mai Kyau)
MCI-AD da AD	12 (17.4%)	57 (82.6%)
MCI-VaD da VaD	38 (50.0%)	38 (50.0%)

Lambar mutuntawa da kashi dari na jimlar marasa lafiya a kowane aji an bambanta su ta nau'in ganewar asali na farko (AD ko VaD).

statistics

Kwatanta halayen mahalarta da sauran fasalulluka na ƙididdigewa tsakanin ƙananan bayanan bayanai don kowane dabarar rarrabuwar samfuri (don hasashen lafiyar fahimi na MoCA da tsananin cutar) an yi ta ta amfani da yaren shirye-shiryen Python (sigar 2.7.1) [17]. An ƙaddamar da bambance-bambancen aikin ƙirar da farko ta amfani da abu ɗaya- ko biyu (kamar yadda ya dace) ANOVA tare da tazarar amincewar 95% da gwajin babban bambanci na gaskiya (HSD) na Tukey don kwatanta ma'anar aikin. An yi wannan gwajin bambance-bambance tsakanin wasan kwaikwayo na ƙirar ta amfani da haɗin Python da R (version 3.5.1) [18]. Mun yi amfani da wannan hanya (duk da haka, ba zato ba tsammani ba ta fi mafi kyau ba) kawai a matsayin taimako na heuristic a wannan. farkon mataki don kwatancen aikin samfur na farko a cikin tsammanin yuwuwar aikace-aikacen asibiti. Daga nan mun yi amfani da gwajin sa hannun Bayesian ta amfani da rarrabawar baya don tantance yuwuwar bambance-bambancen aikin ƙirar [19]. Don waɗannan nazarin, mun yi amfani da tazara -0.01, 0.01, yana nuna cewa idan ƙungiyoyi biyu suna da bambancin aiki na kasa da 0.01, an dauke su iri ɗaya (a cikin yanki na daidaitaccen aiki), ko kuma in ba haka ba sun bambanta (wanda ya fi kyau fiye da). dayan). Don yin kwatancen Bayesian na ƙididdiga da ƙididdige waɗannan yuwuwar, mun yi amfani da ɗakin karatu na baycomp (version 1.0.2) don Python 3.6.4.

Hasashen samfuri

Mun gina nau'ikan tsinkaya ta amfani da jimillar bambance-bambancen tsare-tsare guda goma don yin hasashen (rabe) sakamakon gwajin MoCA na kowane majiyyaci ko tsananin ganewar asibiti. An yi amfani da duk masu koyo kuma an gina samfuran ta amfani da dandalin software na buɗe tushen Weka [20]. Don bincikenmu na farko, mun yi amfani da algorithms na koyo guda 10 da aka saba amfani da su: 5-Maƙwabta mafi kusa, nau'ikan bishiyar yanke shawara ta C4.5, Sake dawo da Logistic, Multilayer Perceptron, Naïve Bayes, nau'ikan Random Forest, Radial Basis Function Network, da Taimakon Vector. Inji. An bayyana mahimman halayen waɗannan algorithms a wani wuri [21] (duba rataye). An zaɓi waɗannan ne saboda suna wakiltar nau'ikan ɗalibai iri-iri kuma saboda mun nuna nasara ta amfani da su a cikin nazarin da suka gabata akan bayanai iri ɗaya. An zaɓi saitunan hyper-parameter daga bincikenmu na baya yana nuna su da ƙarfi akan bayanai daban-daban [22]. Dangane da sakamakon binciken mu na farko ta yin amfani da tsarin haɗe-haɗe iri ɗaya tare da fasali gama-gari waɗanda aka yi amfani da su a cikin cikakken bincike, mun gano xaliban guda uku waɗanda suka samar da aiki mai ƙarfi a duk faɗin rarrabuwa: Rikicin Logistic, Naïve Bayes, da Injin Tallafin Vector.

Tsare-tsare da ƙididdiga aikin ƙira

Don duk ƙirar ƙira (gami da bincike na farko), kowane samfurin an gina shi ta amfani da ingancin giciye mai ninki 10, kuma an auna aikin ƙirar ta amfani da Wurin Ƙarƙashin Ƙarƙashin Ƙarƙashin Ƙarƙashin Mai karɓa (AUC). Tabbatarwa ya fara ne tare da rarraba kowane tsarin ƙirar ƙira guda 10 ba tare da izini ba zuwa kashi 10 daidai gwargwado (folds), ta amfani da tara daga cikin waɗannan sassa daban-daban don horar da ƙirar da sauran ɓangaren don gwaji. An maimaita wannan hanya sau 10, ta yin amfani da wani yanki daban-daban kamar yadda aka saita gwajin a kowane juzu'i. Daga nan aka haɗa sakamakon don ƙididdige sakamakon/aiki na ƙarshe. Ga kowane ɗalibi / haɗewar saitin bayanai, an maimaita wannan gabaɗayan tsari sau 10 tare da raba bayanan daban-daban kowane lokaci. Wannan mataki na ƙarshe ya rage son zuciya, tabbatar da maimaitawa, kuma ya taimaka wajen tantance aikin ƙirar gabaɗaya. Gabaɗaya (don makin MoCA da tsare-tsaren rarrabuwar cututtuka a hade), ƙila 6,600 an gina su. Wannan ya haɗa da ƙira 1,800 waɗanda ba a tace su ba (tsari 6 da aka yi amfani da su akan dataset × 3 xalibai × 10 gudu × folds 10 = 1,800 samfuri) da 4,800 tace samfuri (tsarin ƙirar ƙira 4 da aka yi amfani da ma'ajin bayanai × 3 masu koyo × 4 dabarun zaɓin zaɓi × 10 gudu × 10 ninka = 4,800 samfuri).

Zaɓin fasalin

Don samfuran da aka tace, zaɓin fasalin (ta amfani da hanyoyin martaba huɗu) an yi su a cikin tabbatarwar giciye. Ga kowane nau'i na 10, kamar yadda kashi 10 na daban na bayanan bayanan shine bayanan gwaji, kawai manyan abubuwan da aka zaɓa guda huɗu don kowane saitin horo (watau sauran ninki tara, ko sauran kashi 90% na duk bayanan) an yi amfani da su. don gina samfurori. Ba mu iya tabbatar da waɗanne siffofi guda huɗu aka yi amfani da su a cikin kowane samfuri ba, saboda ba a adana wannan bayanin ko samar da su a cikin dandalin ƙirar da muka yi amfani da su (Weka). Koyaya, idan aka ba da daidaito a cikin zaɓin farkon mu na manyan fasalulluka lokacin da aka yi amfani da masu daraja ga duk bayanan da aka haɗa da kuma kamanceceniya ta gaba a cikin wasan kwaikwayo, waɗannan fasalulluka iri ɗaya (shekarun, shekarun ilimi, MTx-% C, da ma'anar MTx-RT). ) ƙila su ne mafi rinjaye na sama huɗu da aka yi amfani da su tare da zaɓin fasalin a cikin tsarin tabbatar da giciye.

Sakamakon

Halayen ƙididdiga na mahalarta (ciki har da maki MoCA da ma'aunin aikin MemTrax) na ma'auni na bayanai daban-daban na kowane dabarar rarraba ƙirar ƙira don hasashen lafiyar fahimi na MoCA (na al'ada da MCI) da tsananin ganewar asali (mai laushi da mai tsanani) an nuna su a cikin Table 3.

Tebur 3

Halayen mahalarta, makin MoCA, da aikin MemTrax don kowane dabarar rarraba samfurin

Dabarun Rabewa	Shekaru	Ilimi	MoCA Daidaita	MoCA Ba a daidaita ba	MTx-% C	MTx-RT
MoCA Category	Shekaru 61.9 (13.1)	Shekaru 9.6 (4.6)	19.2 (6.5)	18.4 (6.7)	74.8% (15.0)	1.4s (0.3)
Tsananin Ganewa	Shekaru 65.6 (12.1)	Shekaru 8.6 (4.4)	16.7 (6.2)	15.8 (6.3)	68.3% (13.8)	1.5s (0.3)

Ƙimar da aka nuna (ma'ana, SD) waɗanda aka bambanta ta hanyar ƙirar ƙira ta dabarun ƙira sune wakilcin haɗaɗɗun bayanan da aka yi amfani da su don hasashen lafiyar fahimi na MoCA (MCI da na al'ada) da ƙananan bayanan XL da aka yi amfani da su kawai don hasashen tsananin cutar (mai laushi da mai tsanani).

Ga kowane haɗakar maki MoCA (daidaitacce/mara daidaita) da ƙofa (26/23), an sami bambancin ƙididdiga (p = 0.000) a cikin kowane kwatancen nau'i-nau'i (lafiya ta al'ada ta al'ada da MCI) don shekaru, ilimi, da aikin MemTrax (MTx-% C da MTx-RT). Kowane ƙananan bayanan majiyyaci a cikin nau'in MCI na kowane haɗin kai yana kan matsakaicin kimanin shekaru 9 zuwa 15, ya ruwaito kimanin shekaru biyar na ilimi, kuma yana da ƙarancin aikin MemTrax ga ma'auni biyu.

Sakamakon aikin ƙirar ƙididdiga na ƙididdige ƙima na MoCA ta amfani da manyan xaliban uku, Regression Logistic, Naïve Bayes, and Support Vector Machine, an nuna su a cikin Tebur 4. Waɗannan ukun an zaɓi su ne bisa mafi tsayin daka da cikakken aikin ɗalibi a duk nau'ikan nau'ikan iri daban-daban. An yi amfani da su zuwa bayanan bayanan don duk tsarin ƙirar ƙira. Don saitin bayanan da ba a tace ba da ƙirar ƙira, kowane ɗayan bayanan da ke cikin Tebura 4 yana nuna aikin ƙirar bisa ga ma'anar AUC da aka samo daga ƙira 100 (10 runs × folds 10) wanda aka gina don kowane ɗalibi / haɗin tsarin ƙirar ƙira, tare da mafi girma bi da bi. mai koyan aikin da aka nuna a cikin m. Ganin cewa ƙirar bayanan da aka tace, sakamakon da aka ruwaito a cikin Tebura 4 yana nuna matsakaicin matsakaicin wasan kwaikwayon samfuri daga ƙira 400 ga kowane ɗalibi ta amfani da kowane ɗayan hanyoyin martaba (hanyoyin ƙira 4 × 10 yana gudana × 10 ninka).

Tebur 4

Sakamakon rarrabuwar makin MoCA Dichotomous (AUC; 0.0-1.0) ga kowane ɗayan manyan xaliban guda uku don duk tsarin ƙirar ƙira.

An Yi Amfani da Saitin Siffar	Makin MoCA	Matakin yankewa	Juyin Juya Harkar	Naïve Bayes	Tallafa Kayan Injin
Ba a tace ba (fasali 10)	Daidaitawa	23	0.8862	0.8913	0.8695
		26	0.8971	0.9221	0.9161
	Ba a daidaita ba	23	0.9103	0.9085	0.8995
		26	0.8834	0.9153	0.8994
Tace (fasali 4)	Daidaitawa	23	0.8929	0.8954	0.8948
		26	0.9188	0.9247	0.9201
	Ba a daidaita ba	23	0.9135	0.9134	0.9122
		26	0.9159	0.9236	0.9177

Yin amfani da bambance-bambancen saitin fasali, maki MoCA, da madaidaicin maki na MoCA, ana nuna mafi girman aiki ga kowane tsarin ƙirar ƙira a cikin m (ba lallai ba ne a ƙididdiga daban-daban fiye da sauran waɗanda ba a ciki ba m don samfurin da ya dace).

Kwatanta xalibai a cikin dukkan nau'ikan makin MoCA da ƙofa (daidaitacce/marasa daidaitawa da 23/26, bi da bi) a cikin haɗaɗɗen bayanan da ba a tace ba (watau ta amfani da fasalulluka guda 10 na gama gari), Naïve Bayes gabaɗaya shine babban koyo tare da gabaɗaya. Rahoton da aka ƙayyade na 0.9093. Idan aka yi la’akari da manyan xaliban uku, gwaje-gwajen sa hannun Bayesian da ke da alaƙa sun nuna cewa yuwuwar (Pr) na Naïve Bayes wanda ya fi karfin Logistic Regression ya kasance 99.9%. Haka kuma, tsakanin Naïve Bayes da Support Vector Machine, yuwuwar kashi 21.0% na daidaitaccen aiki a cikin aikin ɗalibi (saboda haka, yuwuwar 79.0% na Naïve Bayes ya zarce Injin Tallafin Tallafi), haɗe tare da yuwuwar 0.0% na Injin Tallafin Na'urar yana yin mafi kyau, aunawa. yana ƙarfafa fa'idar aiki don Naïve Bayes. Ƙarin kwatancen nau'in makin MoCA a cikin duk masu koyo/fari yana ba da shawarar ɗan fa'idar aiki ta amfani da makin MoCA marasa daidaituwa tare da daidaitawa (0.9027 da 0.8971, bi da bi; Pr (ba a daidaita ba) = 0.988). Hakazalika, kwatancen ƙofa na yanke hukunci a cikin duk xalibai da nau'ikan maki na MoCA sun nuna ƙaramin fa'idar aikin rarrabuwa ta amfani da 26 a matsayin matakin rarrabuwa tsakanin 23 (0.9056 da 0.8942, bi da bi; Pr (26> 23) = 0.999). A }arshe, nazarin aikin rarrabuwar kawuna ga samfura ta yin amfani da ingantaccen sakamako (watau manyan abubuwa huɗu kawai), Naïve Bayes (0.9143) a adadi mai yawa shine ƙwararrun ƙwararrun ƙwararrun ƙwararrun ƙwararrun ma'auni a duk nau'ikan makin MoCA/kofa. Duk da haka, a cikin dukkan fasahohin kima a hade, duk ɗaliban da suka yi fice sun yi irin wannan. Gwaje-gwajen sa hannun Bayesian sun nuna yuwuwar daidaitaccen aiki 100% tsakanin kowane ɗalibin da aka tace. Kamar yadda yake tare da bayanan da ba a tace ba (ta yin amfani da duk abubuwan gama gari guda 10), an sake samun fa'idar aiki don sigar da ba a daidaita ba ta makin MoCA (Pr (ba a daidaita ba> daidaitacce) = 1.000), kazalika da fa'ida iri ɗaya iri ɗaya don ƙimar ƙima na 26 (Pr (26> 23) = 1.000). Musamman ma, matsakaicin aikin kowane ɗayan manyan xaliban uku a duk nau'ikan makin MoCA/kofa ta amfani da manyan abubuwa huɗu kawai ya wuce matsakaicin aikin kowane ɗalibi akan bayanan da ba a tace ba. Ba abin mamaki ba ne, aikin rarrabuwa na samfuran da aka tace (ta yin amfani da manyan abubuwa huɗu masu daraja) gabaɗaya ya kasance mafi girma (0.9119) zuwa samfuran da ba a tace su ba (0.8999), ba tare da la'akari da ƙirar hanyar martabar fasalin da aka kwatanta da waɗancan ƙirar ta amfani da duk 10 gama gari ba. fasali. Ga kowace hanyar zaɓin fasalin, akwai yuwuwar 100% na fa'idar aiki akan samfuran da ba a tace su ba.

Tare da marasa lafiya da aka yi la'akari da su don rarrabuwa mai tsanani na AD, bambance-bambance tsakanin rukuni (MCI-AD da AD) don shekaru (p = 0.004), ilimi (p = 0.028), Makin MoCA wanda aka daidaita / ba a daidaita shi ba (p = 0.000), da MTx-% C (p = 0.008) sun kasance masu mahimmanci; alhali ga MTx-RT bai kasance (p = 0.097). Tare da waɗancan marasa lafiya da aka yi la'akari da su don rarrabuwa mai tsanani na ganewar asali na VaD, bambance-bambance tsakanin rukuni (MCI-VaD da VaD) don makin MoCA da aka daidaita/mara daidaita.p = 0.007) da MTx-% C (p = 0.026) da MTx-RT (p = 0.001) sun kasance masu mahimmanci; amma ga shekaru (p = 0.511) da ilimi (p = 0.157) babu wani muhimmin bambance-bambance tsakanin rukuni.

Sakamakon aikin ƙirar ƙididdiga na ƙididdiga masu tsanani na ganewar asali ta amfani da ɗalibai uku da aka zaɓa a baya, Logistic Regression, Naïve Bayes, da Support Vector Machine, an nuna su a cikin Tebur 5. Yayin da ƙarin xaliban da aka bincika sun nuna ƙwaƙƙwaran wasan kwaikwayo daban-daban tare da ɗaya daga cikin nau'i biyu na ganewar asibiti. , xalibai ukun da muka gano a matsayin mafi kyawu a ƙirarmu ta baya sun ba da mafi daidaiton aiki tare da sabbin tsare-tsaren ƙirar ƙira. Kwatanta masu koyan a kowane nau'in ganewar asali na farko (AD da VaD), babu wani daidaiton rarrabuwar kawuna tsakanin masu koyo na MCI-VaD da VaD, kodayake Na'urar Tallafin Vector gabaɗaya ta yi fice sosai. Hakazalika, babu wani gagarumin bambance-bambance tsakanin masu koyo don MCI-AD da rarrabuwar AD, kodayake Naïve Bayes (NB) yana da ɗan fa'idar aiki fiye da Logistic Regression (LR) kuma kawai ƙarancin ƙarancinsa akan Injin Tallafin Vector, tare da yuwuwar 61.4% ya canza zuwa +41.7%. A cikin duka bayanan biyu, akwai fa'idar aiki gabaɗaya don Injin Tallafin Vector (SVM), tare da Pr (SVM> LR) = 0.819 da Pr (SVM> NB) = 0.934. Ayyukan rarrabuwar mu gabaɗaya a cikin duk xalibai wajen tsinkayar tsananin ganewar asali a cikin ƙananan bayanan XL ya fi kyau a cikin nau'in ganewar asali na VaD da AD.Pr (VAD> AD) = 0.998).

Tebur 5

Dichotomous ganewar asali na asibiti tsanani aikin rarrabuwa (AUC; 0.0-1.0) sakamakon ga kowane ɗayan manyan xaliban guda uku don tsarin ƙirar ƙira.

Tsarin Samfura	Juyin Juya Harkar	Naïve Bayes	Tallafa Kayan Injin
MCI-AD da AD	0.7465	0.7810	0.7443
MCI-VaD da VaD	0.8033	0.8044	0.8338

Ana nuna mafi girman aikin kowane tsarin ƙirar ƙira a ciki m (ba lallai ba ne a ƙididdiga daban-daban fiye da sauran waɗanda ba a ciki ba m).

DISCUSSION

Ganewar farko na canje-canje a lafiyar hankali yana da mahimmanci amfani mai amfani a cikin kula da lafiyar mutum da lafiyar jama'a iri ɗaya. Lallai, shi ma babban fifiko ne a cikin saitunan asibiti don marasa lafiya a duk duniya. Manufar da aka raba ita ce faɗakar da marasa lafiya, masu ba da kulawa, da masu bayarwa da kuma faɗakarwa a baya dacewa da magani mai tsada da kulawa na dogon lokaci ga waɗanda suka fara samun raguwar fahimi. Haɗa rukunin bayanan asibitocinmu guda uku, mun gano ƙwararrun ɗalibai guda uku waɗanda aka fi so (tare da sanannen sanannen -Naïve Bayes) don gina samfuran tsinkaya ta amfani da su. Ma'auni na aikin MemTrax wanda zai iya dogaro da gaske rarrabuwar yanayin lafiyar hankali dichotomously (lafiya ta yau da kullun ko MCI) kamar yadda za a nuna ta jimlar MoCA. Musamman ma, gaba ɗaya aikin rarrabuwa ga duk xaliban uku ya inganta lokacin da samfuranmu suka yi amfani da manyan abubuwa huɗu kawai waɗanda galibi suka ƙunshi waɗannan ma'aunin aikin MemTrax. Haka kuma, mun bayyana ingantaccen yuwuwar amfani da xalibai iri ɗaya da ma'aunin aikin MemTrax a cikin tsarin ƙirar ƙirar tallafi don tantance tsananin nau'i biyu na gano cutar hauka: AD da VaD.

Gwajin ƙwaƙwalwar ajiya shine tsakiyar gano farkon AD [23, 24]. Don haka, yana da kyau cewa MemTrax abin karɓa ne, mai jan hankali, kuma mai sauƙin aiwatarwa akan layi. gwajin gwaji don ƙwaƙwalwar episodic a cikin yawan jama'a [6]. Daidaiton ganewa da lokutan amsawa daga wannan aikin ci gaba na ci gaba yana bayyana musamman a cikin gano farkon da haɓaka lalacewa da kuma rashin lalacewa a cikin tsarin neuroplastic da ke da alaka da ilmantarwa, ƙwaƙwalwa, da fahimta. Wato, samfuran nan waɗanda suka dogara ne akan ma'aunin aikin MemTrax suna da hankali kuma suna da yuwuwar a sauƙaƙe kuma tare da ƙarancin farashi suna bayyana ƙarancin neuropathologic na ilimin halitta yayin matakin asymptomatic na tsaka-tsaki da kyau kafin ƙarin hasarar aiki mai ƙarfi [25]. Ashford et al. yayi nazari sosai akan tsari da halaye na ƙimar ƙwaƙwalwar ƙima da lokacin amsawa a cikin masu amfani da kan layi waɗanda suka shiga da kansu tare da MemTrax [6]. Girmama cewa waɗannan rabe-raben suna da mahimmanci a cikin mafi kyawun ƙirar ƙira da haɓaka ingantaccen aikace-aikacen kulawa da haƙuri mai inganci, ayyana ƙimar da aka dace a asibiti da bayanan bayanan lokacin amsa yana da mahimmanci wajen kafa mahimman tushe mai mahimmanci don amfanin asibiti da bincike. Ƙimar aiki na MemTrax a cikin gwajin AD don raunin fahimi na farko da tallafin bincike na daban yana buƙatar sannan a yi nazari sosai a cikin mahallin yanayin asibiti inda za'a iya la'akari da abubuwan da suka dace da fahimi, da hankali, da kuma iyawar motar da ke shafar aikin gwaji. Kuma don sanar da hangen nesa na ƙwararru da ƙarfafa amfani mai amfani na asibiti, yana da mahimmanci a farko don nuna kwatancen gwajin ƙima na fahimi, duk da cewa na ƙarshe na iya gane ta ta hanyar dabarun gwaji masu wahala, ilimi da hana harshe, da tasirin al'adu [26] . Dangane da wannan, ingantacciyar kwatancen MemTrax a cikin ingancin asibiti zuwa MoCA wanda galibi ana ɗauka azaman ma'aunin masana'antu yana da mahimmanci, musamman lokacin yin la'akari mafi sauƙin amfani da karɓar haƙuri na MemTrax.

Binciken da ya gabata yana kwatanta MemTrax zuwa MoCA yana ba da haske game da dalili da shaidar farko da ke ba da garantin binciken ƙirar mu [8]. Koyaya, wannan kwatancen da aka rigaya ya danganta kawai ma'aunin aikin MemTrax guda biyu da muka bincika tare da matsayi na fahimi kamar yadda MoCA ta ƙaddara da ƙayyadaddun jeri da ƙimar yankewa. Mun zurfafa ƙimancin amfanin asibiti na MemTrax ta hanyar binciko tsarin tsinkaya na tushen ƙirar ƙima wanda zai ba da ƙarin la'akari na keɓaɓɓu na wasu ƙayyadaddun ƙayyadaddun takamaiman majinyata. Ya bambanta da wasu, ba mu sami fa'ida ba a cikin ƙirar ƙira ta amfani da gyaran ilimi (gyara) zuwa makin MoCA ko kuma a cikin sãɓã wa jũna kiwon lafiya wariya MoCA aggregate makin daga asali shawarar 26 zuwa 23 [12, 15]. A haƙiƙa, fa'idar aikin rarrabuwa an fi so ta amfani da makin MoCA da ba a daidaita shi ba da mafi girma kofa.

Mabuɗin mahimmanci a cikin aikin asibiti

Koyon na'ura galibi ana amfani da shi mafi inganci kuma yana da tasiri a ƙirar ƙira lokacin da bayanai ke da yawa kuma masu girma dabam, wato, lokacin da aka sami abubuwan lura da yawa da ɗimbin ɗabi'u masu ƙima (mai ba da gudummawa). Duk da haka, tare da waɗannan bayanan na yanzu, ƙirar da aka tace tare da zaɓin fasali huɗu kawai waɗanda aka yi mafi kyau fiye da waɗanda ke amfani da duk abubuwan gama gari 10. Wannan yana nuna cewa jimillar bayanan asibitocinmu ba su da mafi kyawun fasali na asibiti (masu ƙima) don tantance marasa lafiya ta wannan hanyar. Koyaya, fifikon fasalin fasalin akan ma'aunin aikin MemTrax - MTx-% C da MTx-RT - suna da ƙarfi suna goyan bayan gina ƙirar gaci ga matakin farko a kusa da wannan gwajin wanda ke da sauƙi, mai sauƙin gudanarwa, ƙarancin farashi, da bayyanawa daidai game da Ayyukan ƙwaƙwalwar ajiya, aƙalla a yanzu azaman allo na farko don rabe-raben binaryar halin lafiyar fahimi. Idan aka ba da ƙwanƙwasa koyaushe akan masu samarwa da tsarin kiwon lafiya, hanyoyin gwajin haƙuri da aikace-aikacen asibiti yakamata a haɓaka su yadda ya kamata tare da mai da hankali kan tattarawa, bin diddigin, da ƙira waɗannan halayen haƙuri da ma'aunin gwaji waɗanda suka fi amfani, fa'ida, da kuma tabbatar da inganci a cikin bincike. da goyon bayan gudanarwa na haƙuri.

Tare da ma'aunin ma'aunin MemTrax na maɓalli guda biyu suna tsakiya ga rarrabuwar MCI, babban koyanmu mai ƙwazo (Naïve Bayes) yana da babban aikin tsinkaya a yawancin ƙira (AUC sama da 0.90) tare da madaidaicin-tabbatacce zuwa ƙimar-ƙarya kusa ko ɗan wuce 4 : 1. Aikace-aikacen asibiti na fassara ta amfani da wannan ɗalibin zai kama (daidaita rarraba) ta mafi yawan waɗanda ke da ƙarancin fahimi, yayin da rage farashin da ke hade da kuskuren rarraba wani mai lafiyar fahimi na yau da kullun a matsayin yana da rashi fahimi (tabbatacce ƙarya) ko rasa wannan rarrabuwa a cikin waɗanda suke da ƙarancin fahimi (ƙarya mara kyau). Ko ɗaya daga cikin waɗannan yanayin rarrabuwar kawuna na iya ɗaukar nauyi mara nauyi ga majiyyaci da masu kulawa.

Yayin da a farkon farko da cikakken nazari mun yi amfani da duka xaliban guda goma a kowane tsarin ƙirar ƙira, mun mayar da hankali ga sakamakonmu akan nau'ikan nau'ikan nau'ikan guda uku waɗanda ke nuna ingantaccen aiki mai ƙarfi. Wannan kuma an yi shi ne don haskakawa, bisa waɗannan bayanan, ɗaliban da za su yi tsammanin yin abin dogaro a babban matsayi a aikace-aikacen asibiti mai fa'ida don tantance ƙimar fahimi. Bugu da ƙari, saboda an yi niyya wannan binciken a matsayin bincike na gabatarwa game da amfanin na'ura koyo game da tantance fahimi da waɗannan ƙalubalen na asibiti a kan kari, mun yanke shawarar kiyaye dabarun koyo cikin sauƙi da gamayya, tare da ƙaramar daidaitawa. Muna jin daɗin cewa wannan hanyar na iya iyakance yuwuwar mafi ƙayyadadden ƙayyadaddun damar iya hasashen haƙuri. Hakanan, yayin da horar da samfuran ta amfani da manyan abubuwan kawai (tace hanya) yana ƙara sanar da mu game da waɗannan bayanan (ƙayyadad da gazawar da aka tattara a cikin bayanan da aka tattara da kuma nuna ƙimar haɓakar lokaci da albarkatu na asibiti masu daraja), mun gane cewa bai daɗe ba don taƙaitawa. iyakokin samfuran kuma, sabili da haka, duk (da sauran fasalulluka) yakamata a yi la'akari da su tare da bincike na gaba har sai mun sami ingantaccen bayanin fasali na fifiko waɗanda zasu dace da faɗuwar yawan jama'a. Don haka, mun kuma gane dalla-dalla cewa ƙarin cikakkun bayanai da faɗaɗa wakilci da haɓaka waɗannan da sauran samfuran zasu zama dole kafin haɗa su cikin ingantaccen aikace-aikacen asibiti, musamman don ɗaukar cututtukan cututtukan da ke shafar aikin fahimi wanda zai buƙaci a yi la’akari da shi a cikin ƙarin kimantawa na asibiti.

An ƙara inganta amfanin MemTrax ta hanyar yin ƙirƙira na tsananin cutar dangane da ganewar asibiti daban. Mafi kyawun aikin rarrabuwa gabaɗaya a cikin hasashen tsananin VaD (idan aka kwatanta da AD) bai kasance ba abin mamaki idan aka ba da sifofin bayanin martaba na haƙuri a cikin samfuran musamman ga lafiyar jijiyoyin jini da haɗarin bugun jini, watau hauhawar jini, hyperlipidemia, ciwon sukari, da (ba shakka) tarihin bugun jini. Ko da yake zai kasance mafi kyawawa da dacewa don gudanar da kimantawar asibiti iri ɗaya akan majinyata masu dacewa da lafiyar fahimi na yau da kullun don horar da xaliban da waɗannan ƙarin bayanan da suka haɗa da. Wannan yana da garanti na musamman, kamar yadda MemTrax yayi niyyar amfani da shi da farko don gano matakin farko na gazawar fahimi da kuma bin canjin mutum na gaba. Hakanan yana da kyau cewa mafi kyawuwar rarraba bayanai a cikin tsarin bayanan VaD ya ba da gudummawa a wani bangare ga ingantaccen aikin ƙirar ƙira. Saitin bayanan VaD ya kasance daidai daidai tsakanin azuzuwan biyu, yayin da bayanan AD tare da ƙarancin marasa lafiya na MCI ba su kasance ba. Musamman a cikin ƙananan ma'ajin bayanai, har ma da wasu ƙarin lokuta na iya yin bambanci mai iya aunawa. Dukansu mahanga biyu dalilai ne masu ma'ana waɗanda ke haifar da bambance-bambance a cikin aikin ƙirar ƙira. Koyaya, daidai gwargwado dangana ingantattun ayyuka ga halayen ƙididdige ƙididdiga ko abubuwan da suka dace musamman ga gabatarwar asibiti da aka yi la'akari da su bai kai ba. Duk da haka, wannan labari ya nuna amfanin MemTrax samfurin tsinkayar tsinkaya a cikin rawar tallafin bincike na asibiti yana ba da hangen nesa mai mahimmanci kuma yana tabbatar da neman ƙarin gwaji tare da marasa lafiya a duk faɗin MCI.

Yin aiwatarwa da nuna amfanin MemTrax da waɗannan samfuran a cikin Sin, inda harshe da al'adu suka bambanta sosai da sauran yankuna na kafaffen amfani (misali, Faransa, Netherlands, da Amurka) [7, 8, 27], yana ƙara nuna yuwuwar. don karbuwar duniya da kuma ƙimar asibiti na dandamali na tushen MemTrax. Wannan babban misali ne na ƙoƙarin daidaita bayanai da haɓaka ƙa'idodi na ƙasa da ƙasa masu amfani da ƙirar kayan aiki don tantance fahimi waɗanda aka daidaita kuma a sauƙaƙe don amfani a duk duniya.

Matakai na gaba a ƙirƙira ƙirƙira fahimi da aikace-aikace

Rashin hankali a cikin AD hakika yana faruwa akan ci gaba, ba a cikin matakai masu hankali ko matakai ba [28, 29]. Koyaya, a wannan matakin farkon, burinmu shine mu fara kafa ikonmu na gina ƙirar da ke haɗa MemTrax wanda zai iya bambanta "al'ada" daga "ba al'ada ba". Ƙarin cikakkun bayanai masu ma'ana (misali, hoton kwakwalwa, fasalin kwayoyin halitta, alamomin halitta, cututtuka, da alamun aiki na hadaddun ayyukan da ke buƙatar fahimi sarrafawa) [30] a cikin yankuna daban-daban na duniya, yawan jama'a, da ƙungiyoyin shekaru don horarwa da haɓaka ƙarin ƙwarewa (ciki har da ma'auni mai nauyi) ƙirar injunan koyon injin za su goyi bayan babban matakin haɓaka rarrabuwa, wato, ikon rarraba ƙungiyoyin marasa lafiya tare da. MCI zuwa ƙanana kuma mafi ƙayyadaddun rarrabuwa tare da ci gaba da raguwar fahimi. Bugu da ƙari, bincikar cututtuka na asibiti lokaci guda ga daidaikun mutane a cikin ɓangarorin majinyata na yanki suna da mahimmanci yadda ya kamata horo waɗannan ƙarin haɗaɗɗun samfura masu ƙarfi da tsinkaya. Wannan zai sauƙaƙe ƙarin ƙayyadaddun tsarin sarrafa shari'ar ga waɗanda ke da alaƙa iri ɗaya, tasiri, da ƙarin ƙayyadaddun bayanan fahimi kuma don haka inganta tallafin yanke shawara na asibiti da kulawar haƙuri.

Yawancin bincike na asibiti da suka dace har zuwa yau sun magance marasa lafiya tare da aƙalla ƙarancin rashin lafiya; kuma, a aikace, sau da yawa ana yin yunƙurin sa baki na haƙuri a matakan ci gaba. Koyaya, saboda raguwar fahimi ya fara da kyau kafin a cika ka'idodin asibiti don cutar hauka, ingantaccen allo na tushen MemTrax na farko zai iya ƙarfafa ilimin da ya dace na daidaikun mutane game da cutar da ci gabanta kuma ya ba da hanzari a baya da kuma matakan da suka dace. Don haka, ganowa da wuri zai iya tallafawa abubuwan da suka dace daga motsa jiki, abinci, goyon bayan tunani, da haɓaka zamantakewar jama'a zuwa maganin magunguna da ƙarfafa canje-canjen da ke da alaƙa da haƙuri a cikin ɗabi'a da fahimtar cewa ɗaya ko a cikin jimla na iya ragewa ko yuwuwar dakatar da ci gaban dementia [31, 32] . Bugu da ƙari, tare da tasiri farkon nunawa, Ana iya sa mutane da iyalansu suyi la'akari da gwaje-gwaje na asibiti ko samun shawarwari da sauran tallafin sabis na zamantakewa don taimakawa wajen bayyana tsammanin da niyya da gudanar da ayyukan yau da kullum. Ƙarin ingantaccen aiki da faɗaɗa amfani mai amfani ta waɗannan hanyoyin na iya zama kayan aiki don ragewa ko dakatar da ci gaban MCI, AD, da ADRD ga mutane da yawa.

Lallai, ƙarancin ƙarshen shekarun haƙuri a cikin bincikenmu baya wakiltar yawan al'amuran al'ada tare da AD. Duk da haka, matsakaicin shekarun kowane rukuni da aka yi amfani da su a cikin tsarin ƙirar ƙira wanda ya danganta da ƙimar MoCA/kofa da tsananin ganewar asali (Table 3) yana nuna mafi rinjaye (sama da 80%) kasancewa aƙalla shekaru 50. Wannan rarraba don haka ya dace sosai don haɓakawa, yana tallafawa amfanin waɗannan samfuran a cikin yawan jama'a da ke nuna waɗanda yawanci ke shafa. farkon farawa da rashin lafiyar neurocognitive saboda AD da VaD. Hakanan, shaidun baya-bayan nan da hangen nesa sun jaddada waɗannan abubuwan da aka sani (misali, hauhawar jini, kiba, ciwon sukari, da shan sigari) waɗanda ke iya ba da gudummawa ga mafi girma da wuri. Babban haɗari na jijiyoyin bugun jini da na tsakiyar rayuwa da sakamakon raunin kwakwalwar jijiyoyin bugun jini wanda ke tasowa tare da bayyananniyar sakamako har ma a cikin matasa. manya [33-35]. Saboda haka, mafi kyawun damar dubawa ta farko don ganowa da wuri mataki na kasawar fahimi da kuma ƙaddamar da ingantacciyar rigakafi da dabarun shiga cikin nasarar magance cutar hauka zai fito daga nazarin abubuwan da ke ba da gudummawa da alamomin da suka gabata a cikin bakan shekaru, gami da balagagge na farko da yuwuwar har da ƙuruciya (lura da dacewa da abubuwan kwayoyin halitta kamar apolipoprotein E daga farkon ciki).

A aikace, ingantattun bincike-bincike na asibiti da matakai masu tsada don ɗaukar hoto na ci-gaba, ba da bayanin kwayoyin halitta, da auna ma'auni masu alamar halitta ba koyaushe a shirye suke ba ko ma mai yiwuwa ga masu samarwa da yawa. Don haka, a lokuta da yawa, ƙila a samu rarrabuwa ta farko ta gaba ɗaya daga ƙila ta amfani da wasu ma'auni masu sauƙi waɗanda majiyyaci ya bayar (misali, rahoton kansa. matsalolin ƙwaƙwalwar ajiya, magunguna na yanzu, da iyakokin ayyuka na yau da kullum) da siffofin alƙaluma na kowa [7]. Rijista kamar Jami'ar California Kiwon Lafiya Rijista (https://www.brainhealthregistry.org/) [27] da sauran waɗanda ke da babban fa'ida na alamun da aka ba da rahoton kai, matakan ƙima (misali, barci da fahimtar yau da kullun), magunguna, matsayin lafiya, da tarihi, da ƙarin cikakkun bayanai na alƙaluma za su zama kayan aiki don haɓakawa da tabbatar da aikace-aikacen aikace-aikacen waɗannan ƙarin samfuran farko a cikin asibitin. Bugu da ari, gwaji kamar MemTrax, wanda ya nuna amfani wajen tantance aikin ƙwaƙwalwar ajiya, na iya samar da ƙwaƙƙwaran ƙididdigewa na ilimin cututtukan AD fiye da alamomin nazarin halittu. Ganin cewa ainihin fasalin cututtukan cututtukan AD shine rushewar neuroplasticity da kuma babban asarar synapses, wanda ke bayyana a matsayin episodic. tabarbarewar ƙwaƙwalwar ajiya, ma'aunin da ke tantance ƙwaƙwalwar ajiya na iya zahiri samar da mafi kyawun ƙididdiga na nauyin ƙwayar cuta na AD fiye da alamomin halitta a cikin mai haƙuri mai rai [36].

Tare da duk nau'ikan tsinkaya-ko an haɗa su ta hanyar hadaddun bayanai da haɗaɗɗun bayanai daga fasaha na zamani da ingantaccen fahimtar asibiti a cikin yankuna da yawa ko waɗanda ke iyakance ga ƙarin asali da ingantaccen bayanin halayen bayanan bayanan mara lafiya da ke wanzu-fahimtar fa'idar basirar wucin gadi. koyon injin shine cewa samfuran da suka haifar zasu iya haɗawa da haɓakawa "koyi" daga sabbin bayanai da hangen nesa da aka samar ta hanyar amfani da aikace-aikacen da ke gudana. Bayan canja wurin fasaha mai amfani, kamar yadda ake amfani da samfuran nan (kuma za a haɓaka) kuma ana wadatar da su tare da ƙarin lokuta da bayanan da suka dace (ciki har da marasa lafiya da ke da cututtukan da ke iya haifar da raguwar fahimi mai zuwa), aikin tsinkaya da rarrabuwar lafiyar fahimi za su yi ƙarfi, yana haifar da ƙarin ingantaccen yanke shawara na tallafi mai amfani. Wannan juyin halitta zai kasance cikakke kuma a aikace tare da shigar da MemTrax zuwa al'ada (wanda aka yi niyya ga iyawar da ake da shi) waɗanda masu ba da lafiya za su iya amfani da su a ainihin lokacin a asibitin.

Mahimmanci ga inganci da amfani da samfurin MemTrax don tallafin bincike da kulawar haƙuri ana nema sosai-bayan bayanan tsayi mai ma'ana. Ta hanyar lura da yin rikodin canje-canje masu haɗaka (idan akwai) a cikin matsayi na asibiti a cikin isassun kewayon al'ada ta farkon matakin MCI, ana iya horar da samfuran don ƙimar da ta dace da ci gaba da rarrabuwa kuma ana iya horar da su kuma a gyara su yayin da marasa lafiya suka tsufa kuma ana kula da su. Wato, maimaita amfani na iya taimakawa tare da bin diddigin sauye-sauye masu sauƙi na fahimi, tasirin sa baki, da kiyaye ingantaccen kulawa. Wannan hanya ta dace sosai tare da aikin asibiti da haƙuri da kula da shari'a.

gazawar

Muna godiya da ƙalubalen da ƙima a cikin tattara bayanan asibiti mai tsabta a cikin asibiti mai kulawa / saitin asibiti. Duk da haka, da zai ƙarfafa ƙirar mu idan kundin bayananmu ya haɗa da ƙarin marasa lafiya tare da fasali gama gari. Bugu da ƙari, musamman ga ƙirar ƙirar mu, da ya kasance mafi kyawu da dacewa don gudanar da kima iri ɗaya na asibiti akan majinyata masu dacewa da lafiyar fahimi na yau da kullun don horar da xaliban. Kuma kamar yadda babban aikin rarrabuwa ya nuna ta yin amfani da tsararren bayanai (kawai manyan abubuwan da aka fi so guda huɗu), ƙari gabaɗaya kuma matakan lafiya/masu nuni da ƙila sun inganta yin ƙirar ƙira tare da mafi girman adadin abubuwan gama gari a duk marasa lafiya.

Wasu mahalarta na iya kasancewa tare suna fuskantar wasu cututtuka waɗanda zasu iya haifar da ƙarancin fahimi ko na yau da kullun. Baya ga bayanan bayanan XL inda aka rarraba marasa lafiya kamar yadda suke da AD ko VaD, ba a tattara bayanan haɗin gwiwa / ba da rahoto a cikin tafkin marasa lafiya na YH, kuma babban abin da ya fi dacewa ya ba da rahoton rashin daidaituwa a cikin KM sub-dataset shine ciwon sukari. Yana da ƙima, duk da haka, cewa haɗawa da marasa lafiya a cikin tsarin ƙirar mu tare da cututtuka masu haɗari waɗanda zasu iya faɗakarwa ko ƙara girman ƙarancin fahimi da ƙananan aikin MemTrax zai zama mafi wakilci na yawan majinyata da aka yi niyya na ainihin duniya don wannan ƙarin fahimi na farko na tantancewa. da tsarin tallan kayan kawa. Ci gaba, ingantacciyar ganewar cututtukan cututtukan da ke da yuwuwar yin tasiri ga aikin fahimi yana da fa'ida sosai don haɓaka samfura da sakamakon aikace-aikacen kula da haƙuri.

A ƙarshe, majinyatan ƙananan bayanan bayanan YH da KM sun yi amfani da wayar hannu don ɗaukar gwajin MemTrax, yayin da ƙayyadaddun adadin majinyatan bayanan bayanan XL sun yi amfani da iPad kuma sauran sun yi amfani da wayar hannu. Wannan zai iya gabatar da ƙaramin bambance-bambancen da ke da alaƙa da na'ura a cikin aikin MemTrax don ƙirar ƙira ta MoCA. Duk da haka, bambance-bambance (idan akwai) a cikin MTx-RT, alal misali, tsakanin na'urori na iya zama marasa mahimmanci, musamman tare da kowane ɗan takara ana ba da gwajin "aiki" kafin aikin gwajin da aka yi rikodin. Duk da haka, amfanin waɗannan na'urorin hannu guda biyu na iya yin illa ga kwatanta kai tsaye zuwa da/ko haɗin kai tare da wasu sakamakon MemTrax inda masu amfani suka amsa don maimaita hotuna ta taɓa ma'aunin sararin samaniya a madannai na kwamfuta.

Maɓalli masu mahimmanci akan kayan aikin tsinkaya na MemTrax

• • Samfuran tsinkayar mu masu inganci waɗanda ke tattare da zaɓaɓɓun ma'aunin aikin MemTrax zai iya dogaro da gaske rarrabuwa yanayin lafiyar fahimi (lafiyar fahimi ta al'ada ko MCI) kamar yadda gwajin MoCA ɗin da aka sani ke nunawa.
• • Waɗannan sakamakon suna goyan bayan haɗawa da zaɓaɓɓun ma'aunin aikin MemTrax zuwa aikace-aikacen tantance samfurin tsinkaya don rashin fahimtar matakin farko.
• • Tsarin ƙirar mu kuma ya bayyana yuwuwar yin amfani da aikin MemTrax a aikace-aikace don bambance tsananin cutar hauka.

Wadannan binciken sabon labari sun kafa tabbataccen shaida da ke tallafawa amfanin koyo na injina wajen gina ingantattun samfuran rarrabuwa na tushen MemTrax don tallafin bincike a cikin ingantaccen tsarin kula da shari'ar asibiti da kulawar haƙuri ga daidaikun mutane masu fama da nakasuwa.

ACKNOWLEDGMENTS

Mun fahimci aikin J. Wesson Ashford, Curtis B. Ashford, da abokan aiki don haɓakawa da kuma tabbatar da aikin ci gaba da ƙwarewa da kayan aiki (MemTrax) da aka yi amfani da su a nan kuma muna godiya ga yawancin marasa lafiya da ciwon hauka waɗanda suka ba da gudummawa ga bincike mai mahimmanci. . Muna kuma gode wa Xianbo Zhou da takwarorinsa na SJN Biomed LTD, abokan aikinsa da kuma masu hadin gwiwa a wuraren asibitoci da asibitoci, musamman Drs. M. Luo da M. Zhong, wadanda suka taimaka wajen daukar mahalarta aiki, da tsara jadawalin gwaje-gwaje, da tattarawa, yin rikodi, da sarrafa bayanan gaba-gaba, da kuma mahalarta masu aikin sa kai wadanda suka ba da lokacinsu mai kima, suka kuma himmatu wajen daukar jarrabawar da samar da su. bayanai masu kimar da za mu tantance a cikin wannan binciken. Wannan Binciken da aka goyan bayan wani ɓangare ta MD Kimiyyar Kimiyya Shirin Jami'ar Kiwon Lafiya ta Kunming (Bayar da lambar 2017BS028 zuwa XL) da Shirin Bincike na Sashen Kimiyya da Fasaha na Yunnan (Bayar da lambar 2019FE001 (-222) zuwa XL).

J. Wesson Ashford ya shigar da takardar neman izini don amfani da takamaiman ci gaba mai ƙima da aka bayyana a cikin wannan takarda don gaba ɗaya. gwajin ƙwaƙwalwar ajiya.

MemTrax, LLC kamfani ne mallakin Curtis Ashford, kuma wannan kamfani yana sarrafa abubuwan gwajin ƙwaƙwalwar ajiya tsarin da aka bayyana a cikin wannan takarda.

Ana samun bayanan bayanan marubuta akan layi (https://www.j-alz.com/manuscript-disclosures/19-1340r2).

nassoshi

[1] Ƙungiyar Alzheimer (2016) 2016 Bayanan cutar Alzheimer da adadi. Alzheimers Dement 12, 459-509.
[2] Gresenz CR , Mitchell JM , Marrone J , Federoff HJ (2019) Tasirin matakin farko Alzheimer ta cutar akan sakamakon kudi na gida. Lafiya Econ 29, 18-29.
[3]	Foster NL , Bondi MW , Das R , Foss M , Hershey LA , Koh S , Logan R , Poole C , Shega JW Neurology: Saitin ma'aunin ingancin ƙarancin fahimi. Kwayoyin cuta 2019, 93-705.
[4] Tong T, Thokala P, McMillan B, Ghosh R, Brazier J (2017) Amfanin farashi gwaje-gwajen tantance fahimi don gano ciwon hauka da ƙarancin fahimi a cikin kulawa na farko. Int J Geriatr Likitan tabin hankali 32, 1392–1400.
[5] Ashford JW , Gere E , Bayley PJ (2011) Aunawa ƙwaƙwalwar ajiya a cikin manyan saitunan rukuni ta amfani da ci gaba da gwajin tantancewa. J Alzheimers Dis 27, 885-895.
[6]	Ashford JW, Tarpin-Bernard F, Ashford CB, Ashford MT (2019) Aiki na ci gaba da ganewa na kwamfuta don auna ƙwaƙwalwar episodic. J Alzheimers Dis 69, 385-399.
[7]	Bergeron MF , Landset S , Tarpin-Bernard F , Ashford CB , Khoshgoftaar TM , Ashford JW (2019) Ƙwaƙwalwar ƙwaƙwalwar ajiya a cikin ƙirar koyo na inji don tsinkaya rabe-raben yanayin kiwon lafiya. J Alzheimers Dis 70, 277-286.
[8] van der Hoek MD , Nieuwenhuizen A , Keijer J , Ashford JW (2019) Gwajin MemTrax idan aka kwatanta da kima na fahimi na ƙwararru na ƙarancin fahimi. J Alzheimers Dis 67, 1045-1054.
[9]	Falcone M , Yadav N , Poellabauer C , Flynn P (2013) Yin amfani da keɓaɓɓen sautunan wasali don rarrabuwa na rauni mai rauni mai sauƙi. A cikin 2013 taron kasa da kasa na IEEE akan Acoustics, Magana da Gudanar da Sigina, Vancouver, BC, shafi 7577–7581.
[10]	Dabek F, Caban JJ (2015) Yin amfani da manyan bayanai don yin la'akari da yiwuwar bunkasa yanayin tunanin mutum bayan rikici. Procedia Comput Sci 53, 265-273.
[11]	Climent MT , Pardo J , Munoz-Almaraz FJ , Guerrero MD , Moreno L (2018) Bishiyar yanke shawara don ganowa da wuri na rashin fahimta ta hanyar masana harhada magunguna na al'umma. Gaban Pharmacol 9, 1232.
[12]	Nasreddine ZS , Phillips NA , Bedirian V , Charbonneau S , Whitehead V , Collin I , Cummings JL , Chertkow H (2005) Ƙididdigar Ƙwararrun Ƙwararru na Montreal, MoCA: Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwararren Ƙwaƙwalwa na Ƙwaƙwalwa na Ƙwaƙwalwa na Ƙaƙƙa ) na Ƙaƙa na Ƙarfafawa na Ƙarfafawa . J Am Geriatr Soc 53, 695–699.
[13]	Yu J, Li J, Huang X (2012) Sigar Beijing ta kimanta fahimi na duniya a matsayin ɗan gajeren kayan aikin tantancewa don ƙarancin fahimi: Nazarin tushen al'umma. BMC Ƙwararrun Ƙwararru 12, 156.
[14]	Chen KL. J Am Geriatr Soc 2016, e64-e285.
[15]	Carson N , Leach L , Murphy KJ (2018) Wani sake-bincike na Montreal Cognitive Assessment (MoCA) yanke maki. Int J Geriatr Ƙwararrun Ƙwararru 33, 379-388.
[16]	Ƙungiyar Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Ƙwararrun Amirka, Inc., Washington, DC.
[17]	Python. Python Software Foundation, http://www.python.org, An shiga Nuwamba 15, 2019.
[18]	R Core Group, R: Harshe da muhalli don Ƙididdigar Ƙididdigar R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/, 2018, Shiga Nuwamba 15, 2019.
[19]	Benavoli A , Corani G , Demsar J , Zaffalon M (2017) Lokaci don canji: Koyawa don kwatanta nau'i-nau'i masu yawa ta hanyar nazarin Bayesian. J Mach Koyi Res 18, 1–36.
[20]	Frank E , Hall MA , Witten IH (2016) The WEKA Workbench. A ciki Haƙar Ma'adinan Bayanai: Kayan Aikin Koyon Inji da Dabaru, Frank E, Hall MA, Witten IH, Pal CJ, eds. Morgan Kaufmann https://www.cs.waikato.ac.nz/ml/weka/Witten_et_al_2016_appendix.pdf
[21]	Bergeron MF , Landset S , Maugans TA , Williams VB , Collins CL , Wasserman EB , Khoshgoftaar TM (2019) Koyon na'ura a cikin ƙirar tsarin wasanni na makarantar sakandaren rikice-rikice. Wasannin Wasanni na Med Sci 51, 1362-1371.
[22]	Van Hulse J, Khoshgoftaar TM, Napolitano A (2007) Ra'ayoyin gwaji akan koyo daga bayanan da ba su da daidaituwa. A ciki Abubuwan da aka gabatar na taron kasa da kasa na 24th akan koyan inji, Corvalis, Oregon, Amurka, shafi 935-942.
[23]	Ashford JW, Kolm P. J Gerontol 1989, 44–139.
[24] Ashford JW , Jarvik L (1985) Alzheimer ta cutar: Shin filastik neuron yana haifar da lalatawar neurofibrillary axonal? N Engl J Med 313, 388-389.
[25] Jack CR Jr, Therneau TM, Weigand SD, Wiste HJ, Knopman DS, Vemuri P, Lowe VJ, Mielke MM, Roberts RO, Machulda MM, Graff-Radford J, Jones DT, Schwarz CG, Gunter JL , Sen. , Rocca WA , Petersen RC (2019) Yawaitar nazarin halittu vs asibiti ma'anar abubuwan bakan Alzheimer ta amfani da Cibiyar Kasa kan tsufa-Alzheimer's Binciken Ƙungiya tsarin aiki. JAMA Neurol 76, 1174-1183.
[26] Zhou X , Ashford JW (2019) Ci gaba a cikin kayan aikin tantancewa Alzheimer ta cutar. Shekaru Med 2, 88-93.
[27] Weiner MW. Kiwon Lafiya Rijista: Dandali na tushen intanet don daukar ma'aikata, kimantawa, da kuma sa ido na tsawon lokaci na mahalarta don nazarin ilimin neuroscience. Alzheimers Dement 14, 1063-1076.
[28] Ashford JW , Schmitt FA (2001) Tsarin lokaci-lokaci Alzheimer dementia. Curr Likitan tabin hankali Rep 3, 20-28.
[29]	Li X, Wang X, Su L, Hu X, Han Y (2019) Nazarin dogon lokaci na Sino kan raguwar fahimi (SILCODE): Yarjejeniya don nazarin dogon nazari na dogon lokaci na kasar Sin don haɓaka nau'ikan hasashen haɗarin juzu'i zuwa ƙarancin fahimi ga mutane masu fahimi. raguwa. BMJ Buɗe 9, e028188.
[30] Tarnanas I , Tsolaki A , Wiederhold M , Wiederhold B , Tsolaki M (2015) Canjin ci gaban biomarker na shekaru biyar don Cutar cutar Alzheimer Hasashen: Shin hadadden ayyukan kayan aiki na alamar rayuwa ta yau da kullun na iya cike giɓi? Alzheimers Dement (Amst) 1, 521-532.
[31] McGurran H , Glenn JM , Madero EN , Bott NT (2019) Rigakafin da maganin cutar Alzheimer: Hanyoyin motsa jiki na halitta. J Alzheimers Dis 69, 311-338.
[32] Mendiola-Precoma J , Berumen LC , Padilla K , Garcia-Alcocer G (2016) Magunguna don rigakafi da maganin cutar Alzheimer. Biomed Res Int 2016, 2589276.
[33]	Lane CA , Barnes J , Nicholas JM , Sudre CH , Cash DM , Malone IB , Parker TD , Keshavan A , Buchanan SM , Keuss SE , James SN , Lu K , Murray-Smith H , Wong A , Gordon E , Coath W , Modat M, Thomas D, Richards M, Fox NC, Schott JM (2020) Ƙungiyoyin tsakanin haɗarin jijiyoyi a fadin girma da kuma ilimin kwakwalwa a cikin marigayi rayuwa: Shaida daga ƙungiyar haihuwa ta Biritaniya. JAMA Neurol 77, 175-183.
[34]	Seshadri S (2020) Rigakafin cutar hauka-tunanin fiye da shekaru da akwatunan amyloid. JAMA Neurol 77, 160-161.
[35]	Maillard P , Seshadri S , Beiser A , Himali JJ , Au R , Fletcher E , Carmichael O , Wolf PA , DeCarli C (2012) Hanyoyin hawan jini na systolic akan mutuncin fararen fata a cikin matasa a cikin Nazarin Zuciya na Framingham: Giciye - nazarin sashi. Lancet Neurol 11, 1039-1047.
[36] Fink HA , Linskens EJ , Silverman PC , McCarten JR , Hemmy LS , Ouellette JM , Greer NL , Wilt TJ , Butler M (2020) Daidaitaccen gwajin biomarker don ma'anar neuropathologically Cutar Alzheimer a cikin tsofaffi masu ciwon hauka. Ann Intern Med 172, 669-677.

Haɗin kai: [a] Binciken SIVOTEC, Boca Raton, FL, Amurka | [b] Sashen Kwamfuta da Injiniyan Lantarki da Kimiyyar Kwamfuta, Jami'ar Florida Atlantic University, Boca Raton, FL, Amurka | [c] SJN Biomed LTD, Kunming, Yunnan, China | [d] Cibiyar don Binciken Alzheimer, Cibiyar Nazarin Kiwon Lafiya ta Washington, Washington, DC, Amurka | [e] Sashen Magungunan Gyaran Gwiwa, Asibitin Farko mai alaƙa na Jami'ar Kiwon Lafiya ta Kunming, Kunming, Yunnan, China | [f] Sashen Nazarin Jiki, Asibitin Jama'ar Dehong, Dehong, Yunnan, China | [g] Sashen Nazarin Jiki, Asibitin Farko na Jami'ar Kiwon Lafiyar Kunming, gundumar Wuhua, Kunming, lardin Yunnan, na kasar Sin | [h] Cibiyar Nazarin Cututtuka da Cutar da ke da alaƙa, VA Palo Alto Health Care Tsarin, Palo Alto, CA, Amurka | [i] Ma'aikatar Ilimin Hauka & Kimiyyar Halayyar, Makarantar Magunguna ta Jami'ar Stanford, Palo Alto, CA, Amurka

Sadarwa: [*] Magana zuwa: Michael F. Bergeron, PhD, FACSM, SIVOTEC Analytics, Boca Raton Innovation Campus, 4800 T-Rex Avenue, Suite 315, Boca Raton, FL 33431, Amurka. E-mail: mbergeron@sivotecanalytics.com.; Xiaolei Liu, MD, Sashen Nazarin Jiki, Asibitin Farko na Jami'ar Kiwon Lafiyar Kunming, Titin Xichang 295, gundumar Wuhua, Kunming, lardin Yunnan na 650032, kasar Sin. E-mail: ring@vip.163.com.

Keywords: tsufa, Alzheimer ta cutar, ciwon hauka, duban taro