Respiratory Diseases Diagnosis Using Audio Analysis and Artificial Intelligence: A Systematic Review

Respiratory diseases represent a significant global burden, necessitating efficient diagnostic methods for timely intervention. Digital biomarkers based on audio, acoustics, and sound from the upper and lower respiratory system, as well as the voice, have emerged as valuable indicators of respiratory functionality. Recent advancements in machine learning (ML) algorithms offer promising avenues for the identification and diagnosis of respiratory diseases through the analysis and processing of such audio-based biomarkers. An ever-increasing number of studies employ ML techniques to extract meaningful information from audio biomarkers. Beyond disease identification, these studies explore diverse aspects such as the recognition of cough sounds amidst environmental noise, the analysis of respiratory sounds to detect respiratory symptoms like wheezes and crackles, as well as the analysis of the voice/speech for the evaluation of human voice abnormalities. To provide a more in-depth analysis, this review examines 75 relevant audio analysis studies across three distinct areas of concern based on respiratory diseases' symptoms: (a) cough detection, (b) lower respiratory symptoms identification, and (c) diagnostics from the voice and speech. Furthermore, publicly available datasets commonly utilized in this domain are presented. It is observed that research trends are influenced by the pandemic, with a surge in studies on COVID-19 diagnosis, mobile data acquisition, and remote diagnosis systems.

The crystallographic symmetry test for the correctness of a set of phases

One of the most important problems in the application of direct methods for large structures is to establish reliable consistency criteria for the correctness of a phasing trial. The introduction in the twin variables method [Bethanis, Tzamalis, Hountas, Mishnev & Tsoucaris (2000). Acta Cryst. A56, 105-111] of a new criterion based on the crystallographic symmetry consists of testing the phase extension and refinement algorithm by deliberately sacrificing the space-group-symmetry information in the auxiliary variable set then using its gradual re-appearance as a criterion for correctness. In the present paper, the crystallographic symmetry test has been used in the implementation of the twin algorithm in two different ways: (i) as an overall test throughout the iterations that is likely to reflect the correctness of the phasing procedure for each one of the extension trials in a macromolecular phasing environment; (ii) as a convenient criterion to determine the optimum cycle for freeing the initial phases used by the algorithm for the phase-extension procedure.

Ab initio determination of a crystal structure by means of the Schrödinger equation

It is shown that the use of the Schrödinger equation may lead to the ab initio determination of the positions of the nuclei in a crystal, given a limited number of diffracted-beam intensities. In particular, it is shown that an extremely simplified Schrödinger equation in physical momentum space provides a sufficiently sound theoretical basis to develop an algorithm using diffraction data alone. This algorithm has been tested with a known 41 atom crystalline structure (not including hydrogen atoms) in space group P1. The extracted information is sufficient to determine the positions of all atoms. In addition, theoretical developments relevant to the connection between momentum space in quantum mechanics and diffraction theory have been formulated.

New developments of the TWIN algorithm for phase extension and refinement in disordered supramolecular structures

A new development of the TWIN algorithm is described and used for phase extension/refinement in supramolecular complexes. A small number of phased reflections at low resolution are sufficient for the quasi-automated determination of all atomic coordinates, including disordered atoms.

Upgrading the twin variables algorithm for large structures

Phase extension from lower to higher resolution by using an upgraded TWIN variables algorithm [Hountas & Tsoucaris (1995). Acta Cryst. A51, 754-763] in protein molecules with close to 1,000 non-H atoms is presented. Three points of this procedure are of particular interest. (i) The use of a set of auxiliary variables providing a satisfactory fit for many kinds of constraints: the new algorithm works efficiently despite the extreme 'dilution' of very limited initial phase information into a much larger set of auxiliary variables. (ii) The extension of this auxiliary variables set beyond the resolution of the observed data, which enhances the phase extension in a so-called 'super-resolution' sphere. (iii) The use of the crystallographic symmetry as a new figure of merit and as a reliable test for the correctness of the phase-extension process allows an efficient screening.