Prof. Dr. Jan Steckel, lead researcher of the CORELSA project, would like to thank all people involved. Read his letter here.
COSYS-LAB REMOTE LUNG SOUND ANALYSIS
Large-scale remote lung monitoring of patients with respiratory syndromes
CORELSA (CoSys-Lab Remote Lung Sound Analysis) is an open-source project that tackles the issue that arises when hospitals are dealing with a large number of patients with respiratory syndromes, like during the current COVID-19 pandemic. One of the most important and most frequent examinations are auscultations with stethoscopes, which take a lot of time for the medical staff. Moreover, because of the risk of infection, the use of traditional stethoscopes is becoming increasingly difficult. The CORELSA
System offers a solution to reduce the workload for the physicians and nurses, diminishes the risk of infections, helps to reduce the use of protective clothing and leads to more standardized auscultations.
The CORELSA System consists of auscultation units with three digital stethoscopes, which are placed on the patients’ chest. The CORELSA auscultation units can be built using low cost off-the-shelf components in combination with the open source code and 3D models available on our GIT-repository. Once all components are gathered, the system can be assembled and installed quickly, with only some tweaking steps. The system consists of up to three custom-made digital stethoscopes. In our tutorials we explain how to build these stethoscopes out of small microphones, 3D printed adapters and traditional stethoscopes. The digital stethoscopes can be plugged into the jacks on the side of the auscultation box. Within the box the stethoscopes are connected to USB sound cards, which themselves are connected to the Raspberry Pi. We recommend using IP67 enclosures and connectors (see manual) to make sure the system is waterproof and can be disinfected properly if needed. The auscultation units can then be linked via Ethernet or Wifi to a network so that they can be registered by the framework. This framework has no limitations in terms of number of auscultation units that can be connected.
by physicians from the University Hospital of Antwerp (UZA). The three stethoscopes should be fixed with medical tape (see picture) and eventually also gaze bandage on the desired positions. While we recommend to use three stethoscopes, it is also possible to run our system with only one or two stethoscopes. To avoid any confusion and to be able to trace which measurement is acquired from which stethoscope, the colored tubing matches the color template of the 3D printed adapters and the audio connectors. The whole system is designed to be intuitive to use, to save time for the medical personnel.
We build our digital stethoscopes using Standard Littmann Classic III stethoscopes and Sharkoon SM1 microphones, which are connected via 3D printed connetctors. To ensure the quality of the recordings, it is important to securely attach the stethoscopes to the preferred locations on the patients body. In the CORELSA user interface we suggest eight different locations that were recommended
Once the auscultation units are connected to the network, a central node will handle all traffic and provide an interface that can be used to control the units and acquire auscultations remotely. This webpage-interface lists all connected patients and has several useful diagnostic features that were developed in close collaboration with physicians of the University Hospital of Antwerp (UZA). Devices are automatically registered and a patient name and/or patient ID can be linked to a unit. The user can decide whether periodical measurements are required for continuous monitoring, and/or if they
want to execute a measurement on the spot. Once the measurement is available, the user can listen to the recording of all stethoscopes or view their respective spectrograms. The spectrogram window has a zoom function to investigate certain spectrogram features or artifacts if needed. It is also possible to note the presence or absence of certain diagnostic sound features or write remarks and link them to the recording for future reference.