Machine Learning project about prediction ICU admissions following an operation
We will conduct this research for the company Chipsoft. Chipsoft is an Amsterdam based software company which develops software for the healthcare sector. Chipsoft will deliver us a dataset from which we can derive information about patients during a surgery. The umbrella term PDMS will be used to refer to the data. PDMS stands for Patient Data Management System.
Our data is provided by Chipsoft. The data is a csv file which consists of the following columns: 1. MeetTijd 2. HR (bpm) (Heartbeat) 3. SpO2(%) (Oxygenation) 4. NIBP(mmHg) 5. IcNaOk
In the following format
| MeetTijd | HR(bpm) | SpO2(%) | NIBP(mmHg) | IcNaOk |
|---|---|---|---|---|
| [10:26:49, 10:27:49, 10:28:49, 10:29:50, 10:32... | [83, 90, 85, 89, 87, 85, 89, 84, 89, 84, 82, 9... | [98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98, 9... | [154, 154, 154, 154, 152, 152, 152, 131, 131, ... | 0 |
| [15:06:54, 15:07:54, 15:08:54, 15:09:55, 15:10... | [51, 53, 52, 53, 52, 55, 53, 56, 60, 50, 50, 5... | [92.5, 94, 92, 93, 93, 92, 92, 92, 93, 98, 97,... | [156, 156, 156, 158, 158, 158, 157, 157, 157, ... | 1 |
| [08:31:20, 08:32:21, 08:33:21, 08:34:21, 08:35... | [63, 67, 67, 102, 99, 98, 80, 81, 76, 76, 94, ... | [98, 99, 99, 99, 99, 99, 98, 98, 98, 98, 98, 9... | [160, 160, 160, 145, 145, 145, 174, 174, 174, ... | 0 |
How does the architecture of the Neural Network influence whether someone will end up at the intensive care unit or will have complications following surgery given their PDMS?
1. What is a patient data management system?
2. How does a deep learning machine algorithm work?
1. Can we find correlations in the data that can give us useful insights?
2. Can we provide the sergeants with insights derived from the data during the surgery