BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Unveiling Causal Drivers of Non-Communicable Diseases with Interpr
 etable Models - Sheresh Zahoor
DTSTART:20240524T161500Z
DTEND:20240524T170000Z
UID:TALK217186@talks.cam.ac.uk
CONTACT:Pietro Lio
DESCRIPTION:In healthcare\, where accurate and reliable decision-making is
  paramount\, interpretability is essential. Traditional Machine Learning (
 ML) models have provided valuable insights but often lack transparency in 
 their reasoning\, limiting their effectiveness. The recent surge in ML tec
 hniques across medical fields such as radiology\, cardiology\, mental heal
 th\, and pathology holds great promise. These techniques can improve diagn
 ostic accuracy\, enhance workflow efficiency\, minimise medical errors\, a
 nd ultimately improve public health outcomes. However\, the "black-box" na
 ture of many ML algorithms raises significant concerns about interpretabil
 ity. The lack of transparency in these models' decision-making processes o
 ften prevents clear explanations for their predictions\, which undermines 
 trust and hinders their integration into clinical practice. This issue has
  led to a growing movement towards interpretable models in healthcare\, sh
 ifting away from traditional approaches. Probabilistic graphical models (P
 GMs)\, particularly Causal Bayesian Networks (CBNs)\, are emerging as fron
 t-runners in interpretable models for healthcare. CBNs offer a framework f
 or representing causal relationships between variables. This fosters a dee
 per understanding of the mechanisms influencing healthcare outcomes. Integ
 rating domain knowledge and expert clinical insights empowers CBNs to capt
 ure more accurate causal relationships between risk factors and health out
 comes.  This enriched model provides a more realistic understanding of hea
 lthcare phenomena\, as it goes beyond simply identifying correlations and 
 unveils the underlying causal drivers. By prioritizing interpretable model
 s like CBNs\, we empower healthcare professionals to make informed decisio
 ns and develop improved preventative strategies. This ultimately leads to 
 superior patient outcomes. Our research prioritises Non-Communicable Disea
 ses (NCDs) like diabetes and cardiovascular diseases (CVD) due to their si
 gnificant public health burden. These chronic illnesses are often preventa
 ble through lifestyle modifications\, highlighting the importance of ident
 ifying key modifiable risk factors. To achieve this\, we conducted an exte
 nsive analysis utilizing various structural learning algorithms. This anal
 ysis helped us identify causal pathways among potential risk factors affec
 ting the progression of these diseases.  Based on these pathways\, we deve
 loped novel CBNs that represent the identified causal relationships. These
  CBNs offer valuable insights into the progression and prevention of NCDs\
 , empowering healthcare professionals with a powerful tool to combat these
  diseases at their root cause.
LOCATION:Lecture Theatre 2\, Computer Laboratory\, William Gates Building
END:VEVENT
END:VCALENDAR