Validation of the Artificial Intelligence-Based Predictive Optimal Trees in Emergency Surgery Risk (POTTER) Calculator in Emergency General Surgery and Emergency Laparotomy Patients

Hechi, Majed W. El; Maurer, Lydia R.; Levine, Jordan; Zhuo, Ying Daisy; Moheb, Mohamad El; Velmahos, George C.; Dunn, Jack; Bertsimas, Dimitris; Kaafarani, Haytham M.A.

doi:10.1016/j.jamcollsurg.2021.02.009

Cited by 38 publications

(25 citation statements)

References 28 publications

(27 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Decision trees (DTs) find a wide range of practical uses [190,189,108,179,26,29,28,69,25,27,30,32,33,115,72,34,37,31,142,145,52,24,174,171,36,67,127,35,19]. Moreover, DTs are the most visible example of a collection of machine learning (ML) models that have recently been advocated as essential for high-risk applications [160].…”

Section: Introductionmentioning

confidence: 99%

On Tackling Explanation Redundancy in Decision Trees

Izza,

Ignatiev,

Marques-Silva

2022

Preprint

View full text Add to dashboard Cite

Decision trees (DTs) epitomize the ideal of interpretability of machine learning (ML) models. The interpretability of decision trees motivates explainability approaches by socalled intrinsic interpretability, and it is at the core of recent proposals for applying interpretable ML models in high-risk applications. The belief in DT interpretability is justified by the fact that explanations for DT predictions are generally expected to be succinct. Indeed, in the case of DTs, explanations correspond to DT paths. Since decision trees are ideally shallow, and so paths contain far fewer features than the total number of features, explanations in DTs are expected to be succinct, and hence interpretable. This paper offers both theoretical and experimental arguments demonstrating that, as long as interpretability of decision trees equates with succinctness of explanations, then decision trees ought not be deemed interpretable. The paper introduces logically rigorous path explanations and path explanation redundancy, and proves that there exist functions for which decision trees must exhibit paths with explanation redundancy that is arbitrarily larger than the actual path explanation. The paper also proves that only a very restricted class of functions can be represented with DTs that exhibit no explanation redundancy. In addition, the paper includes experimental results substantiating that path explanation redundancy is observed ubiquitously in decision trees, including those obtained using different tree learning algorithms, but also in a wide range of publicly available decision trees. The paper also proposes polynomial-time algorithms for eliminating path explanation redundancy, which in practice require negligible time to compute. Thus, these algorithms serve to indirectly attain irreducible, and so succinct, explanations for decision trees. Furthermore, the paper includes novel results related with duality and enumeration of explanations, based on using SAT solvers as witness-producing NP-oracles.

show abstract

Section: Introductionmentioning

confidence: 99%

On Tackling Explanation Redundancy in Decision Trees

Izza,

Ignatiev,

Marques-Silva

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The morbidity C statistics was similarly the highest (0.8414) 73 . Recently, this same score was validated in emergency laparotomy patients 74 …”

Section: Methodology and Outcomesmentioning

confidence: 83%

Value in acute care surgery, part 2: Defining and measuring quality outcomes

Ross¹,

Wandling²,

Bruns³

et al. 2022

J Trauma Acute Care Surg

View full text Add to dashboard Cite

show abstract

“…In the past, other risk scoring systems have been used in the context of TBI patients, including the Charlson Comorbidity Index (CCI) and the APACHE II score [36][37][38][39][40]. Even risk calculators such as Predictive OpTimal Trees in Emergency Surgery Risk (POTTER) could be considered a possible alternative [41,42]. However, each of these indexes have their limitations.…”

Section: Discussionmentioning

confidence: 99%

Mortality risk stratification in isolated severe traumatic brain injury using the revised cardiac risk index

Forssten

Bass

Scheufler

et al. 2021

Eur J Trauma Emerg Surg

View full text Add to dashboard Cite

Purpose Traumatic brain injury (TBI) continues to be a significant cause of mortality and morbidity worldwide. As cardiovascular events are among the most common extracranial causes of death after a severe TBI, the Revised Cardiac Risk Index (RCRI) could potentially aid in the risk stratification of this patient population. This investigation aimed to determine the association between the RCRI and in-hospital deaths among isolated severe TBI patients. Methods All adult patients registered in the TQIP database between 2013 and 2017 who suffered an isolated severe TBI, defined as a head AIS ≥ 3 with an AIS ≤ 1 in all other body regions, were included. Patients were excluded if they had a head AIS of 6. The association between different RCRI scores (0, 1, 2, 3, ≥ 4) and in-hospital mortality was analyzed using a Poisson regression model with robust standard errors while adjusting for potential confounders, with RCRI 0 as the reference. Results 259,399 patients met the study’s inclusion criteria. RCRI 2 was associated with a 6% increase in mortality risk [adjusted IRR (95% CI) 1.06 (1.01–1.12), p = 0.027], RCRI 3 was associated with a 17% increased risk of mortality [adjusted IRR (95% CI) 1.17 (1.05–1.31), p = 0.004], and RCRI ≥ 4 was associated with a 46% increased risk of in-hospital mortality [adjusted IRR(95% CI) 1.46 (1.11–1.90), p = 0.006], compared to RCRI 0. Conclusion An elevated RCRI ≥ 2 is significantly associated with an increased risk of in-hospital mortality among patients with an isolated severe traumatic brain injury. The simplicity and bedside applicability of the index makes it an attractive choice for risk stratification in this patient population.

show abstract

Validation of the Artificial Intelligence-Based Predictive Optimal Trees in Emergency Surgery Risk (POTTER) Calculator in Emergency General Surgery and Emergency Laparotomy Patients

Cited by 38 publications

References 28 publications

On Tackling Explanation Redundancy in Decision Trees

On Tackling Explanation Redundancy in Decision Trees

Value in acute care surgery, part 2: Defining and measuring quality outcomes

Mortality risk stratification in isolated severe traumatic brain injury using the revised cardiac risk index

Contact Info

Product

Resources

About