Robert’s Working Blog

Working on the risk factors and finsihed:

1. Layout

2. Transfer Data

 

Today, debug the code for building new data set and finally, it works. But, the equation for 7 blood parameters should get rid of exp(). As the data is too large in the program.  It means that Frank’s “equations - viz log.xls” is the final excel equation used in my program.

1. finish risk factor data transformation

2. meeting with Guillermo for early grant and had three tasks to do

3. attend Bruce’s three hours course

4. programming on loading risk factors data in 

1. Finish modifying Frank’s code and can transform into “.map” style

2. Finish Creating a new data structure of “Shen.map”

3. Extracting useful risk factors from database and convert them into readable numerical format 

1. Extract useful data field from Access database to excel file
2. Read Excel files into Matlab
3. Create a new data structure for both blood/ultrasound parameter and risk factors
4. Transfer the raw data into “.map” format.

 

Reading code and prepare for programming to transfer raw data into matlab format

1. Reading materials from Julia 

2. Looking at the database 

3. Prepare to extract new data out 

1. Start on looking at the database

2. prepare to export into excel format

3. Transfer them into matlab readable format.

Trip to Newcastle:

 

I make a presentation in the morning and my feedback from this presentation is that:
1. Roger has some concern about our approach to calculate the “average trajectory”. He points out that this trajectory is similar to “growing birth” trajectory but it can’t represent the average term birth.
2. Andrew (one of doctors from John Hunter Hospital) suggests that we should firstly concentrate on “term” birth research and Roger agrees with Andrew’s suggestion. He further points out that “term birth” patient should get rid of patients who are “induced” term birth.
3. Roger is quite happy with our “trajectory analysis”.
 
After that, we have a small group meeting about preterm birth with Frank, Julia, Maria and two other ladies. This meeting is mainly solved my data problem.

1. I’ve got a new database which has 515 patients with about 30 Preterm birth patients. Obviously, there is some problem for this database (for example, there is no twins and if one mother goes to the hospital for the second baby, it will recognize her as two different women). Roger said there will be some change on the database and the new database will be available within next two-three months.
2. Roger suggests that we should focus on “trajectory analysis”. Maybe, we can start with the “term birth” research first.

 
I met Frank to discuss the program to transfer the raw data into matlab readable format. Also I told him about my progress. He may start to work from this weekend.
 
I met Julia and discuss the equation for specific parameter (i.e. CRH has two parameters: á1 and á2). As I have a little bit concern about whether this equation is the best. She seems also have some concern about this. She showed me that she tried CRH=a+bt² and plot the CRH. Unfortunately, this approach still can’t separate the “term” and “pre-term” group. But I do glad to know that the rate of CRH trajectory has some exciting results, so I was thinking maybe we should try this first. I asked her to send me some of her progress report and maybe I can know more about details.
 
I also ask Roger about building “preterm birth” library and he gives me some paper and one of his new journals about preterm birth.
 

 

Meeting with boss:

1. Clear the possible approach way

2. Clear the task for this month/this year

3. The CI approach needs to filter some data and plot the distribution

4. Calculate the difference between patient’s trajecotory and average trajectory will identify the peak time