How to unblock that health data in your EHR? Blockchain. (UK)

The solution to that huge pile of patient-generated data, blocked and stymied in those non-interoperable EHRs [TTA 15 Mar], may be a system based on blockchain. DeepMind, Alphabet’s AI ‘skunk works’, is building a tool that it calls Verifiable Data Audit. It will be tested first in UK hospitals with which DeepMind is already working, including London’s Royal Free Hospital. What VDA will do is use cryptographic math to keep an accurate record of data used in the past to see exactly who is using health-care records, and for what purpose. When data is used, it generates a code based on all past activity. Any alteration to one part of the data alters the others and is quick to spot.

The UK test results will be interesting because, according to the MIT Technology Review article, patient records are considered to be highly fragmented. Another issue that DeepMind had in the UK was the NHS oversharing data with it for other projects, such as AI systems to diagnose eye disease, early warning signs of illness, and machine-learning approaches to guide cancer treatment. The VDA approach would, ironically, create an audit trail of that data. Another reason why we may be moving from Data Despare to Hope. Hat tip to contributor Sarianne Gruber of RCM Answers.

Is telemedicine attractive to hypochondriacs?

An article in MIT Technology Review takes a sideways look at telemedicine and asks if telemedicine is providing an easy route for people suffering from excessive anxiety about their health. The author, Christina Farr, suggests that the ease of contacting a doctor using telemedicine services in comparison to having to visit a doctor’s office and the ability use either insurance or direct payments makes these services more attractive to hypochondriacs (lately called those with somatic symptom disorder).
Views on the subject are quoted from the chief medical affairs officer at MDLive, Deborah Mulligan, and a board member of Doctor on Demand, Bob Kocher. While the first is able to relate an anecdote where a case of excessive anxiety disorder was identified and successfully referred to cognitive behavioral therapy, the latter says he isn’t aware of any patients with health anxiety regularly using the Doctor on Demand app.

Read the full article here.

IBM automates diagnostic image analysis

“Most smart software in use today specialises on one type of data, be that interpreting text or guessing at the content [grow_thumb image=”” thumb_width=”150″ /]of photos. Software in development at IBM has to do all that at once. It’s in training to become a radiologist’s assistant” writes Tom Simonite in MIT Technology Review.

According to Simonite, the IBM software, named Avicenna, analyses diagnostic images like CT scans and the associated data such as a patient’s medical record and suggests possible diagnoses.

An example quoted was the case of a 28-year old with shortness of breath whose pulmonary angiogram images and medical history were analysed by the software. Using a family history which showed a tendency to form blood clots the software diagnosed a pulmonary embolism which was the same diagnosis an independent radiologist reached.

Simonite reports that IBM have thus far used annonymised data and are now working on commercialising the software although an independent researcher is quoted as saying that the accuracy needs to be increased before it would be a useful diagnostic tool.

Read the full article here.