The Clarity- and Virtue-guided Algorithms (CAVA) Lab at Boston Children's Hospital / Harvard Medical School is seeking a post-doctoral research fellow to advance the interpretability and fairness of machine learning (ML) models deployed in critical healthcare settings.  The fellow will join a multi-disciplinary team of computer scientists, informaticists, clinicians, engineers and bioethicists to develop and assess clinical prediction algorithms and advance our understanding of the behavior of machine learning models deployed in health settings. The fellow will help us think critically about how machine learning methods affect clinical practice and outcomes; in particular, 1) the conditions under which ML provides or fails to provide insight into disease pathologies, and 2) the conditions under which ML exacerbates or mitigates treatment and outcome disparities between patient subgroups. 

Prediction models are an increasingly important technology in the digital health landscape, and can produce large-scale changes in  health care via their interactions with patients, clinicians, and hospital operations. This postdoctoral fellowship provides an opportunity to study these issues more deeply in order to improve our ability to diagnose and intervene in a more trustworthy and equitable way. 

The fellowship includes an academic appointment at Harvard Medical School, as well as a hospital appointment at Boston Children’s Hospital. This position provides an excellent opportunity for the Research Fellow to work within a multidisciplinary research team to explore advanced areas in health information technology. CHIP is home to 20 faculty working at the forefront of research areas extending beyond clinical prediction algorithms to domains like clinical NLP, digital epidemiology, clinical genomics, and app ecosystems for health records. CHIP and the CAVA Lab value diversity and believe that it is essential to our research goals. We therefore strongly encourage candidates from underrepresented groups to apply.  

Admissions

The position is available immediately and is renewable annually.

Qualifications
  • PhD degree in computer science, information science, biomedical informatics, data mining, engineering, applied mathematics, or a closely related field.
  • A track record of high-quality research that demonstrates the ability to independently identify important research topics and carry out experiments. 
  • Candidates with strong experience in machine learning, preferably both in the assessment of ML algorithms in data science applications and in the development of novel methods. 
  • Experience and familiarity with the machine learning literature on interpretability and fairness. 
  • Experience working with large, heterogeneous data collections, especially electronic health records, multi-omics data, or other health data.
  • Programming experience in Python, R, and/or C++.
  • Experience with collaborative software development (revision control, continuous integration, etc) strongly preferred 
  • Experience with open science practices (preprints, reproducible workflows, etc) strongly preferred
  • Strong written and oral communication skills.
  • Ability to work both independently and as a team player.
How to apply

Interested candidates should email a CV, three letters of reference, and a sample publication to Dr. William La Cava, PI Clarity and Virtue-guided Algorithms Lab: william.lacava@childrens.harvard.edu.

Publications

Keloth VK, Banda JM, Gurley M, Heider PM, Kennedy G, Liu H, Liu F, Miller T, Natarajan K, V Patterson O, Peng Y, Raja K, Reeves RM, Rouhizadeh M, Shi J, Wang X, Wang Y, Wei WQ, Williams AE, Zhang R, Belenkaya R, Reich C, Blacketer C, Ryan P, Hripcsak G, Elhadad N, Xu H. Representing and Utilizing Clinical Textual Data for Real World Studies: An OHDSI Approach. Journal of biomedical informatics 2023.

Toce MS, Michelson KA, Hudgins JD, Olson KL, Monuteaux MC, Bourgeois FT. Association of prescription drug monitoring programs with benzodiazepine prescription dispensation and overdose in adolescents and young adults. Clinical toxicology (Philadelphia, Pa.) 2023.

Brown T, de Salazar Munoz PM, Bhatia A, Bunda B, Williams EK, Bor D, Miller JS, Mohareb A, Thierauf J, Yang W, Villalba J, Naranbai V, Garcia Beltran W, Miller TE, Kress D, Stelljes K, Johnson K, Larremore D, Lennerz J, Iafrate AJ, Balsari S, Buckee C, Grad Y. Geographically skewed recruitment and COVID-19 seroprevalence estimates: a cross-sectional serosurveillance study and mathematical modelling analysis. BMJ open 2023.

El-Hayek C, Barzegar S, Faux N, Doyle K, Pillai P, Mutch SJ, Vaisey A, Ward R, Sanci L, Dunn AG, Hellard ME, Hocking JS, Verspoor K, Boyle DI. An evaluation of existing text de-identification tools for use with patient progress notes from Australian general practice. International journal of medical informatics 2023.

Patik I, Redhu NS, Eran A, Bao B, Nandy A, Tang Y, El Sayed S, Shen Z, Glickman J, Fox JG, Snapper SB, Horwitz BH. The IL-10 receptor inhibits cell extrinsic signals necessary for STAT1-dependent macrophage accumulation during colitis. Mucosal immunology 2023.