Ua hōʻike ʻo AEYE Health i nā hopena mai kāna hoʻokolohua FDA-clinical pivotal no ka ʻike autonomous o ka retinopathy maʻi maʻi maʻi. ʻO ka haʻawina ka mea mua o kāna ʻano e loiloi inā hiki i ka polokalamu AI ke ʻike pololei i ka retinopathy maʻi maʻi maʻi me ka hoʻohana ʻana i hoʻokahi kiʻi i kēlā me kēia maka, i loaʻa mai ka pākaukau a i ʻole ka pahu kiʻi retinal lima.
Hiki i ka hoʻohana ʻana i hoʻokahi kiʻi mai kēlā me kēia maka ke maʻalahi i ke kaʻina hana diagnostic a hōʻemi i ka manawa nānā.
ʻO nā hopena AI i ʻike ʻia no kēlā me kēia ʻōnaehana kamera:
- Topcon NW-400 (paʻi kiʻi paʻi kiʻi): 93.0% sensitivity, 91.4% kiko'ī, >99% kiʻi.
- Optomed Aurora (paʻi kiʻi paʻa lima): 91.9% sensitivity, 93.6% kiko'ī, >99% imageability
- Ua hoʻohana ʻia kahi kiʻi hoʻokahi no kēlā me kēia maka no ka pākaukau a me nā pahu paʻa lima.
Aia he 35M ma'i ma'i ma'i ma US a ma luna o 420M ma ka honua holo'oko'a e pilikia ana i ka ho'omohala 'ana i ka ma'i diabetic retinopathy e koi ana i ka nānā 'ana i ka makahiki. ʻO ka hoʻomaʻamaʻa mua ʻana a me ka hana ʻana he kī nui no ka pale ʻana i ka nalo. Ma luna o ka hapakolu o ka poʻe maʻi maʻi ma mua o 40 mau makahiki e hoʻomohala i ka retinopathy diabetic, akā naʻe he 15%-50% wale nō o nā maʻi e hoʻopaʻa ʻia.
ʻO ka kānana kūʻokoʻa i ka hoʻonohonoho mālama mua e koi ana i hoʻokahi kiʻi wale nō i kēlā me kēia maka i loaʻa mai kahi pahupaʻi lima kūʻai kūpono e hiki ke hoʻomaikaʻi i ka mālama ʻana i nā protocol screening a i ka hopena e pale i ka makapō i ka maʻi maʻi maʻi i ka pilikia o ka nalo ʻana.
Ke ʻimi nei ʻo AEYE Health i ka hoʻomaʻemaʻe FDA no kāna hopena screening autonomous no ka retinopathy maʻi maʻi.
Eia kekahi, ke hoʻomākaukau nei ka hui e hoʻolaha i kāna mau hopena i ka ʻike ʻana i ka glaucoma me ka hoʻohana ʻana i nā kiʻi kālā kālā.
He aha e lawe ʻia mai kēia ʻatikala:
- ʻO ka kānana kūʻokoʻa i ka hoʻonohonoho mālama mua e koi ana i hoʻokahi kiʻi wale nō i kēlā me kēia maka i loaʻa mai kahi pahupaʻi lima kūʻai kūpono e hiki ke hoʻomaikaʻi i ka mālama ʻana i nā protocol screening a i ka hopena e pale i ka makapō i ka maʻi maʻi maʻi i ka pilikia o ka nalo ʻana.
- The study was the first of its kind to evaluate whether AI software can accurately detect more-than-mild diabetic retinopathy using a single image per eye, obtained from either a desktop or handheld retinal camera.
- There are 35M diabetics in the US and over 420M worldwide who are at risk of developing diabetic retinopathy requiring an annual screening.
