Software as a Medical Device (SaMD) is emerging as a new smart healthcare application, but what counts most for software developers to successfully tap into the medical AI domain is that they must know first what problems their solutions can address for the healthcare industry, according to Mishal Patel, head of health informatics at AstraZeneca (AZ), a UK-based biopharmaceutical company.
Patel said the value of innovative healthcare software products lies in their ability to solve problems seen in the medical treatment process, such as those seen in consumables management and analysis of medical records. Image or text algorithms can also be applied to help hospitals more accurately monitor the volumes of drugs and intravenous drips used and the life of surgery instruments in use, thus allowing hospitals to make timely replenishments, maintenances and replacements.
Patel said the healthcare industry hopes AI can help analyze patients' medical histories to enhance development of new drugs, but drug development is highly complicated. Not only better clinical trials must be designed, but traits of the new drugs and the conditions of the patients should also be more precisely monitored by incorporating tech elements.
He said the accuracy and accessibility of data is also crucial in building effective algorithms, stressing the concept of "FAIR" (findable, available, interpretable and researchable) and the need to unify data formats.