← Back to Research Papers

Can neural networks model the human perception of geometric shapes?

Authors: Pajot M, Morfoisse T, Sablé-Meyer M, Lakretz Y, Dehaene S
Journal: PLoS computational biology
mental health psychology open access

Abstract

Myotonic dystrophy type 1 (DM1) is an autosomal dominant neuromuscular disease caused by a CTG repeat expansion within the gene. DM1 is characterized by progressive muscle weakness and myotonia, accompanied by metabolic manifestations including insulin resistance and dyslipidemia. In addition, overweight and obesity are common. Body weight issues in this patient population likely reflect disease-related factors. First, muscle wasting is inherent to DM1 and will continue with disease progression. Second, motor impairments and diminished motivation or fatigue will decrease physical activity. This might lead to increased cardiovascular morbidity and reduced quality of life. In addition, insulin resistance and altered substrate metabolism may further contribute to overweight and obesity. Central nervous system involvement may also affect motivation and daily functioning, which could indirectly influence lifestyle behaviours, including dietary habits. While a mismatch between energy uptake and energy expenditure can contribute to overweight and obesity, studies also report underweight as a clinically relevant feature, with cachexia observed in approximately 19.7% of patients. Nutritional management can support a healthy body weight; an important component of nutritional management is the estimation of total daily energy expenditure (TEE) derived from basal metabolic rate (BMR) and physical activity level (PAL). BMR, the main component of TEE, is largely determined by fat-free mass (FFM) and can be measured using indirect calorimetry. In clinical practice, BMR is commonly estimated using predictive equations such as the WHO, original Harris–Benedict (1918), revised Harris Benedict (2023) or Mifflin-St Jeor equation. These equations rely on anthropometric variables, such as height and weight, and other physiological determinants of BMR, such as age and sex. Thereafter, PAL should be estimated, which is generally selected from standardized tables based on history taking by a nutritional expert.