Implications of conscious AI in primary healthcare

The conversation about consciousness of artificial intelligence (AI) is an ongoing topic since 1950s. Despite the numerous applications of AI identified in healthcare and primary healthcare, little is known about how a conscious AI would reshape its use in this domain. While there is a wide range of ideas as to whether AI can or cannot possess consciousness, a prevailing theme in all arguments is uncertainty. Given this uncertainty and the high stakes associated with the use of AI in primary healthcare, it is imperative to be prepared for all scenarios including conscious AI systems being used for medical diagnosis, shared decision-making and resource management in the future. This commentary serves as an overview of some of the pertinent evidence supporting the use of AI in primary healthcare and proposes ideas as to how consciousnesses of AI can support or further complicate these applications. Given the scarcity of evidence on the association between consciousness of AI and its current state of use in primary healthcare, our commentary identifies some directions for future research in this area including assessing patients', healthcare workers' and policy-makers' attitudes towards consciousness of AI systems in primary healthcare settings.

A novel immune cell signature predicts pathological complete response to neoadjuvant chemotherapy in triple negative breast cancer patients in the Q-CROC3 trial

e12614

Background: Tumor infiltrating lymphocytes (TILs) have been associated with good prognosis and response to neoadjuvant chemotherapy. Several reports have shown that the heterogeneity of tumor infiltrating immune cells affects the response to chemotherapy, with for example low levels of FOXP3 expressing T cells associated with good prognosis and pathological complete response (pCR) to chemotherapy. Methods: We examined different immune cell markers on 52 pre-chemotherapy biopsy specimens obtained from triple negative breast cancer patients undergoing neo-adjuvant chemotherapy from the Q-CROC-03 trial. Slides were stained for CD8, CD3,PD-1, PDL-1, FOXP-3 and Granzyme B using multi-colour immunohistochemistry and automated cell counting of stroma and epithelial counts was conducted using the Vectra/inForm image analysis platform. We had total of 39 variables for analysis and we performed Penalized logistic regression for variable selection. Results: Nine variables were found statistically significant to predict response to chemotherapy, PD1+ stroma counts being the one with the highest probability of association with response. A tree algorithm was then used on all 9 variables to identify the best variable and threshold combination to identify patients who respond to chemotherapy. We separated our cohort in test (25% of samples n = 13) and training (75% of samples n = 39) sets for this analysis. Restricting the tree depth to 2 variables for clinical interpretability identified the combination of average counts of stromal PD1+ and average density of stromal FOXP3+ as predictors of chemo response (accuracy 0.82). Both stromal average PD1+ counts and average stromal FOXP3+ density positively correlated with the levels of TILS. Conclusions: Combining FOXP3 and PD1 protein expression in the stroma of pre-treatment biopsies of triple negative breast cancers receiving neoadjuvant chemotherapy is highly predictive of pCR.