Reaching the Unreached with AI

Artificial  Intelligence  (AI)  can  be  considered  an  option  to  help  in  the  preliminary diagnosis  of  skin  conditions  where  there  is  no  reach  of  speciality  doctors  such  as Dermatologists.  The  mission  of  “creating  a  way  to  reach  the  unreached”,  a  study presented at Dermacon International 2019 conference, Bangalore.

Dr. Dheeraj N. Rao resident dermatologist D. Y. Patil University – School of Medicine under the guidance of Dr Sharmila Patil – Prof & HOD Dermatology D Y Patil Medical College presented a study at DERMACON INTERNATIONAL 2019  – INDIA, being hosted by the vibrant  team of Dermatologists of IADVL Karnataka. It was the 47th National  Conference   of   Indian  Association   of   Dermatologists,  Venereologists  & Leprologists  and  the  first  ever  DERMACON with  an  international outreach  –  an “Indian mission with a global vision”, held between 17th to 20th January 2019.

The   results??   Dr   Dheeraj   mentioned   in   the   study   that   “Though   AI   needs   an improvement, most of the Non-Dermatology Doctors were outperformed by the AI app Tibot.ai on diagnosing the skin condition,”.

The basic principle of Artificial  Intelligence in the diagnosis of skin condition can be understood by taking an  example  of  a  trainee  dermatologist.  The  accuracy  of  the  trainee  is  enhanced  by looking at multiple cases with similar patterns. The AI works on the same principle, the Deep Neural Network of AI takes an image, converts it into numerical forms and generate  from  its  trained  model  a  prediction  of  the  diagnosis.  The  AI  enabled application, Tibot, has the capacity to learn depending on the amount of data ( history & clinical images ) fed into it.

Using  this  AI  application  (app)  a  structured  study  was  formulated  following  the institutions   ethical   guidelines   that   took   place   at   multiple   places   comprising Dermatology OPD of a Tertiary Care Hospital and private clinics in rural and semi- urban settings, with an objective to analyze the accuracy of an Artificial Intelligence tool in diagnosing skin conditions in real-world Dermatological patients.

The study Involved the following Doctors – participants

  •     BHMS Doctor
  •     BAMS Doctor
  •     MBBS Doctor
  •     Medical Officer at Govt Hospital, MBBS
  •     Resident Dermatologist at DY Patil
  •     Dermatologist Professor at DY Patil
  •     AI app Tibot

The study sample had a total of 398 real world dermatology cases which were analysed by  each  doctors  mentioned and  also  exposed  to  the  AI  application.  Majority  of  the cases were dominated by the Fungal infections (Tinea) which was followed by Eczema, Alopecia,   Acne,   Pigmentation   Disorder,   Psoriasis,   Infestations,   Bacterial   Skin Infection, Viral skin infection, Immunological and others.

When the diagnosis of different doctors who participated in the study was compared it was found that all the local doctors could diagnose the case within the range of min 47.00% to max 57.50% of accuracy, which was still less than the AI app whereas the accuracy of Resident Dermatologist was found to be 90.30% as compared to the gold standard of 100.00% by the Dermatologist professor

Artificial  Intelligence  - eDigital Facts
Figure 1 Comparison of Accuracy of Doctors with AI

The study ended with a conclusion that though the accuracy of the current version of the app is still less than that of the Resident Dermatologist but it performed better than most of the Non-Dermatologist doctors in the semi-urban and rural area

Though in present scenario with current version Artificial  Intelligence appears to lack diagnostic accuracy as  compared  to  a  dermatologist  however  in  the  future  AI  shall  have  a  significant bearing on helping a dermatologist in cases of diagnostic dilemmas.

It is also considered that this AI enabled app will be useful as Education resource, as a assistance in clinical diagnosis especially in an institution with no dermatologist and to a non-dermatologist doctor who is attempting to triage the skin lesions. It will be useful  to  prevent  the  wrong  diagnosis  in  such  situations  and  thereby  initiating  the treatment in the right direction

With a belief that Artificial  Intelligence will revolutionize the standard of care in dermatology Dr Dheeraj mentioned that this technology could potentially shift the full body screening to the primary   care  level,   improving  earlier   detection  of   skin   cancers  and   increasing appropriate referrals. Hence one should embrace AI and learn to integrate it into our patient care paradigms rather than fearing the futuristic robo-dermatologist he added.

About  D  Y  Patil  Hospital

D  Y  Patil  Hospital  or  D.Y.P.H.R.C.  is  a  charitable  hospital  in  Navi  Mumbai,  India.  It was  founded  in  2004  by  His  Excellency  Padmashri  Dnyandeo  Yashwantrao  Patil.  It has 1500 beds dedicated to  charity, a 100-bed ICU facility, 21  operation theatres, a

24×7 charitable casualty and trauma centre.

About  Polyfins  Technology  Inc

Polyfins Technology Inc is a healthcare start up based in California using Artificial

Intelligence to help users manage their skin conditions. Website: https://tibot.ai/

email: info@polyfins.com

arshad@polyfins.com

Phone: +1 858 401 2332

Facebook Comments